GUT MAY 2011

Contents Commentaries Colon 565 The benefits of diagnostic ERCP in autoimmune pancreatitis 648 Flagellin administra...

Author: NONE

15 downloads 2184 Views 15MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Contents

Commentaries

Colon

565 The benefits of diagnostic ERCP in autoimmune pancreatitis

648 Flagellin administration protects gut mucosal tissue from irradiation-induced apoptosis via MKP-7 activity

M M Lerch, J Mayerle

Cover caption: Pancreatic acinar cell. Coloured transmission electron micrograph (TEM) of a slice through several enzyme-secreting acinar cells from a pancreas. The red circles are secretory granules packed with digestive enzymes ready for export to the small intestine. The four round blue and yellow structures are cell nuclei. Each cell is filled with a network of densely folded membrane, called rough endoplasmic reticulum, here coloured light yellow. The membrane’s surface is covered in small dots, called ribosomes. These are protein-manufacturing sites where various digestive enzymes are produced and secreted by the cell. Magnification 31200 at 634.5cm size. Cover credit: SCIENCE PHOTO LIBRARY.

Journal of the British Society of Gastroenterology Impact Factor: 9.35 Editor

Professor Emad El-Omar (UK) Deputy Editor (Hepatology)

Alexander Gerbes (Germany) Deputy Editor (Luminal Gastroenterology)

William Grady (USA) Associate Editors

Ramon Bataller (Spain) Guy Boeckxstaens (Belguim) Laurence Egan (Ireland) Thomas Gress (Germany) Ernst Kuipers (The Netherlands) Julian Panes (Spain) Herbert Tilg (Austria) Fabien Zoulim (France) Editorial Office

Gut BMJ Journals BMA House, Tavistock Square London, WC1H 9JR, UK T: +44 (0)20 7383 6394 F: +44 (0)20 7383 6668 E: [email protected]

ISSN: 0017-5749 (print) ISSN: 1468-3288 (online) Disclaimer: Gut is owned and published by BMJ Publishing Group Ltd, a wholly owned subsidiary of the British Medical Association and the British Society of Gastroenterology. The owners grant editorial freedom to the Editor of Gut. Gut follows guidelines on editorial independence produced by the World Association of Medical Editors and the code on good publication practice of the Committee on Publication Ethics. Gut is intended for medical professionals and is provided without warranty, express or implied. Statements in the Journal are the responsibility of their authors and advertisers and not authors’ institutions, the BMJ Publishing Group, the British Society of Gastroenterology or the BMA unless otherwise specified or determined by law. Acceptance of advertising does not imply endorsement. To the fullest extent permitted by law, the BMJ Publishing Group shall not be liable for any loss, injury or damage resulting from the use of Gut or any information in it whether based on contract, tort or otherwise. Readers are advised to verify any information they choose to rely on. Copyright Ó 2011 BMJ Publishing Group and British Society of Gastroenterology. All rights reserved; no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without prior permission Gut is published by BMJ Publishing Group Ltd, typeset by TNQ Books & Journals, Chennai, India and printed in the UK on acid-free paper from sustainable forests by Latimer Trend & Co Ltd, Plymouth, UK Gut (ISSN 0017-5749) is published monthly by BMJ Publishing Group and distributed in the USA by Mercury International Ltd. Periodicals postage paid at Rahway, NJ. POSTMASTER: send address changes to Gut, Mercury International Ltd, 365 Blair Road, Avenel, NJ, 07001, USA.

Volume 60 Issue 5 | GUT May 2011

567 Long-term proton pump inhibitor administration, H pylori and gastric cancer: lessons from the gerbil J G Fox, E J Kuipers

569 Breaking down haem attenuates acute pancreatitis: a new treatment option? J Mayerle, M Sendler, M M Lerch

Guidelines 571 Guidelines for the management of inflammatory bowel disease in adults C Mowat, A Cole, A Windsor, T Ahmad, I Arnott, R Driscoll, S Mitton, T Orchard, M Rutter, L Younge, C Lees, G-t Ho, J Satsangi, S Bloom, On behalf of the IBD Section of the British Society of Gastroenterology

Oesophagus 608 Roles of Kru¨ppel-like factor 4 in oesophageal epithelial cells in Barrett’s epithelium development H Kazumori, S Ishihara, Y Takahashi, Y Amano, Y Kinoshita

Upper GI cancer 618 Prospective study of serum cysteine levels and oesophageal and gastric cancers in China G Murphy, J-H Fan, S D Mark, S M Dawsey, J Selhub, J Wang, P R Taylor, Y-L Qiao, C C Abnet

Helicobacter pylori

R M Jones, V M Sloane, H Wu, L Luo, A Kumar, M V Kumar, A T Gewirtz, A S Neish

658 Accuracy of computed tomographic colonography in a nationwide multicentre trial, and its relation to radiologist expertise D Heresbach, M Djabbari, F Riou, C Marcus, A Le Sidaner, M A Pierredon-Foulogne, T Ponchon, M Boudiaf, J A Seyrig, H Laumonier, D Luet, M Giraud-Cohen, A L Pelletier, A Charachon, F Ramaholimihaso, P Bouillet, M Veyrac, S Ficarelli, K Vahedi, J Keruhel, H Lamouliatte, C Ridereau-Zins, Y Bouhnik, M Tissier, B Diris, A M Zagdanski, J M Josselin, S Hamonic, Y Gandon

Pancreas 666 Endoscopic retrograde pancreatography criteria to diagnose autoimmune pancreatitis: an international multicentre study A Sugumar, M J Levy, T Kamisawa, G J M Webster, M-H Kim, F Enders, Z Amin, T H Baron, M H Chapman, N I Church, J E Clain, N Egawa, G J Johnson, K Okazaki, R K Pearson, S P Pereira, B T Petersen, S Read, R P Sah, N S Sandanayake, N Takahashi, M D Topazian, K Uchida, S S Vege, S T Chari

671 Panhematin provides a therapeutic benefit in experimental pancreatitis A Habtezion, R Kwan, E Akhtar, S P Wanaski, S D Collins, R J Wong, D K Stevenson, E C Butcher, M B Omary

624 Long-term proton pump inhibitor administration worsens atrophic corpus gastritis and promotes adenocarcinoma development in Mongolian gerbils infected with Helicobacter pylori T Hagiwara, K-i Mukaisho, T Nakayama, H Sugihara, T Hattori MORE CONTENTS <

Inflammatory bowel disease 631 Dysbiosis of the faecal microbiota in patients with Crohn’s disease and their unaffected relatives M Joossens, G Huys, M Cnockaert, V De Preter, K Verbeke, P Rutgeerts, P Vandamme, S Vermeire

Neurogastroenterology 638 Neuronal stimulation with 5-hydroxytryptamine 4 receptor induces anti-inflammatory actions via a7nACh receptors on muscularis macrophages associated with postoperative ileus Y Tsuchida, F Hatao, M Fujisawa, T Murata, M Kaminishi, Y Seto, M Hori, H Ozaki

Contents

Volume 60 Issue 5 | GUT May 2011

Hepatology

Recent advances in basic science

680 Insulin resistance and low-density apolipoprotein B-associated lipoviral particles in hepatitis C virus genotype 1 infection

710 P16 reactivation induces anoikis and exhibits antitumour potency by downregulating Akt/survivin signalling in hepatocellular carcinoma cells

S H Bridge, D A Sheridan, D J Felmlee, S U Nielsen, H C Thomas, S D Taylor-Robinson, R D G Neely, G L Toms, M F Bassendine

688 Community and personal risk factors for hepatitis C virus infection: a survey of 23 820 residents in Taiwan in 1991e2 M-H Lee, H-I Yang, C-L Jen, S-N Lu, S-H Yeh, C-J Liu, S-L You, C-A Sun, L-Y Wang, W J Chen, C-J Chen, for the R.E.V.E.A.L.eHCV Study Group

H Hu, Z Li, J Chen, D Wang, J Ma, W Wang, J Li, H Wu, L Li, M Wu, Q Qian, J Chen, C Su

Recent advances in clinical practice 722 Chronic gastrointestinal ischaemia: shifting paradigms P B F Mensink, L M G Moons, E J Kuipers

JournalScan 695 Nanoliposomal ceramide prevents in vivo growth of hepatocellular carcinoma H R S Tagaram, N A DiVittore, B M Barth, J M Kaiser, D Avella, E T Kimchi, Y Jiang, H C Isom, M Kester, K F Staveley-O’Carroll

702 Working Party proposal for a revised classification system of renal dysfunction in patients with cirrhosis F Wong, M K Nadim, J A Kellum, F Salerno, R Bellomo, A Gerbes, P Angeli, R Moreau, A Davenport, R Jalan, C Ronco, Y Genyk, V Arroyo

738 GI Highlights from the literature

PostScript 739 Letters

Editor’s quiz: GI snapshot 630 A prematurely terminated journey S A Müller-Lissner, E Fuhrmann

657 A case of dyspepsia and abdominal fullness J D Thomas, T M Monaghan, J James

Journal of the British Society of Gastroenterology, a registered charity

Editor Professor Emad El-Omar (UK) Deputy Editor (Hepatology) Alexander Gerbes (Germany) Deputy Editor (Luminal Gastroenterology) William Grady (USA) Associate Editors Ramon Bataller (Spain) Guy Boeckxstaens (Belgium) Laurence Egan (Ireland) Thomas Gress (Germany) Ernst Kuipers (The Netherlands) Julian Panes (Spain) Herbert Tilg (Austria) Fabien Zoulim (France) Specialist Editors Willem de Vos (The Netherlands) Kentaro Sugano (Japan) Education Editor Guruprasad Aithal (UK) Guidelines Editor Simon Greenfield (UK) JournalScan Editors Guruprasad Aithal (UK) Alex Ford (UK) Contributors Neeraj Bhala (UK) Kofi Oppong (UK) Alastair Windsor (UK)

Aims and Scope: Gut is a leading international journal in gastroenterology and hepatology with an established reputation for publishing first class clinical research of the alimentary tract, the liver, biliary tree and pancreas likely to impact on clinical practice within the foreseeable future. Gut delivers up-to-date, authoritative, clinically oriented coverage of all areas in gastroenterology. Regular features include articles by leading authorities, commentaries on published papers, recent advances in basic science and clinical practice, images illustrating important clinical messages (GI snapshots) and letters. The Journal has an authoritative global Editorial Board and a growing international readership. Contact Details

Editorial Board D H Adams (UK) D C Baumgart (Germany) A Benesic (Germany) M Bernardi (Italy) U Beuers (Netherlands) A J Bredenoord (Netherlands) M Camilleri (USA) A Chak (USA) Y Chowers (Israel) J Claria (Spain) J-F Colombel (France) S Danese (Italy) E N De Toni (Germany) P Desreumaux (France) P Ferenci (Austria) S J Forbes (UK) C Gasche (Austria) K Geboes (Belgium) S Ghosh (Canada) D Grundy (UK) S Harrison (USA) C Hellerbrand (Germany) L A Houghton (UK) J A Jankowski (UK) A Kaser (Austria) M M Lerch (Germany) P M Lynch (USA) P Marteau (France) D F Martin (UK)

J V Mayerle (Germany) K E L McColl (UK) S Moller (Denmark) G Monteleone (Italy) D J Mutimer (UK) J M Neuberger (UK) J Panes (Spain) M P Peppelenbosch (Netherlands) M Pinzani (Italy) D Mark Pritchard (UK) G Rogler (Switzerland) F Salerno (Italy) A K Saluja (USA) M Sans (Spain) J Satsangi (UK) M Schemann (Germany) J Schoelmerich (Germany) T Seufferlein (Germany) M Simren (Sweden) J D Soderholm (Sweden) J J Y Sung (Hong Kong) S P L Travis (UK) R Tutuian (Switzerland) G Van Assche (Belgium) G R van den Brink (Netherlands) J R F Walters (UK) A J M Watson (UK) H Wedemeyer (Germany) Editor, BMJ

Book Review Editor Ray J Playford (UK) Statistical Advisor Caroline Sabin (UK) Guidelines for Authors and Reviewers Full instructions are available online at http://gut.bmj.com/ifora. Articles must be submitted electronically http://submit-gut.bmj.com. Authors retain copyright but are required to grant Gut an exclusive licence to publish http://gut.bmj.com/ifora/ licence.dtl

GUT Online Archive The full back archive of Gut is now available from 1960 at www.gutjnl.com. Users have FREE access to all articles, older than 12 months*, right back to Volume 1, Issue 1. The online archives were digitised with the help of the US National Library of Medicine. Personal subscribers must sign in to access this content. Users not covered by an institutional or personal subscription will have to complete a one-time registration that grants access to all archive content, across all our titles. *Articles that are 12 months old or less require a subscription to access them.

Subscription Information Gut is published monthly; subscribers receive all supplements

Institutional Rates 2011

Personal Rates 2011

Print £600; US$1170; 810

Print (includes online access at no additional cost)

Online Site licences are priced on FTE basis and allow access by the whole institution. Print is available at deeply discounted rates for online subscribers; details available online at http://group.bmj.com/ group/subs-sales/subscriptions or contact the Subscription Manager in the UK (see above right).

Online only £132; US$258; 179

£237; US$463; 320

ISSN 0017-5749 (print); 1468-3288 (online)

Personal print or online only and institutional print subscriptions may be purchased online at http://group.bmj.com/group/subs-sales/subscriptions (payment by Visa/Mastercard only). Residents of some EC countries must pay VAT; for details, call us or visit www.bmj.com/subscriptions/ vatandpaymentinfo.dtl

Editorial Office Gut, BMJ Journals BMA House, Tavistock Square London, WC1H 9JR, UK T: +44 (0)20 7383 6394 F: +44 (0)20 7383 6668 E: [email protected] Permissions http://journals.bmj.com/misc/permissions.dtl Supplement Enquiries T: +44 (0)20 7383 6057 F: +44 (0)20 7383 6668 E: [email protected] Subscriptions For all subscription enquiries and orders T: +44 (0)20 7383 6270 F: +44 (0)20 7383 6402 E: [email protected] Display Advertising Sales Nick Gray T: +44 (0)20 7383 6386 F: +44 (0)20 7383 6556 E: [email protected] http://group.bmj.com/group/advertising Online Advertising Sales Marc Clifford T: +44 (0) 20 7383 6161 F: +44 (0) 20 7383 6556 E: [email protected] http://group.bmj.com/group/advertising Author Reprints Reprints Administrator T: +44 (0)150 251 5161 F: +44 (0)207 554 6185 E: [email protected] Commercial Reprints (except USA & Canada) Nadia Gurney-Randall T: +44 (0)20 8445 5825 F: +44 (0)20 8445 5870 M: 07866 262 344 E: [email protected] Commercial Reprints (USA & Canada) Marsha Fogler T: +1 800 482 1450 (toll free in the USA) T: +1 856 489 4446 (outside the USA) F: +1 856 489 4449 E: [email protected] British Society of Gastroenterology 3 St Andrew’s Place, Regents Park, London NW1 4LB, UK T: +44 (0) 20 7935 2815 F: +44 (0) 20 7487 3734 E: [email protected] www.bsg.org.uk

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Digest

Highlights from this issue doi:10.1136/gut.2011.242479

Emad El-Omar, Alexander Gerbes and William Grady, Editor and Deputy Editors

New insights into the factors that British society of cause Barrett’s oesophagus gastroenterology guidelines on The mechanisms responsible for the inflammatory bowel disease

transformation of the normal oesophageal squamous epithelium into the metaplastic intestinal epithelium that characterises Barrett’s oesophagus are poorly understood, but are essential to know in order to develop better ways for preventing Barrett’s oesophagus. Prior work by the authors of this study and by others has shown that a transcription factor called CDX2 is one of the mechanisms involved in the formation of Barrett’s oesophagus. Kazumori and colleagues now show that Kruppel-Like Factor 4 (KLF4), a key protein that regulates the differentiation of the epithelium of the gastrointestinal tract, is a second transcription factor strongly expressed in Barrett’s epithelium that is induced by bile acids. Importantly, they have shown that KLF4 and CDX2 cooperate to induce each other’s expression and that KLF4 induces the expression of the mucin protein MUC2, which is involved in intestinal metaplasia. Moreover, they also show that the inﬂammation related signalling pathway, NF-kB, is also involved in this process. This study advances our understanding of the transcriptional network related to KLF4 and CDX2 that affects the development of Barrett’s oesophagus. KLF4 and CDX2 are possible molecular targets for the prevention of Barrett’s oesophagus, and therapies directed at these proteins may yield a much-needed medical treatment for Barrett’s oesophagus (see page 608).

Model of how KLF4 and the CDX2 transcription factors regulate Barrett’s oesophagus. Gut May 2011 Vol 60 No 5

This issue proudly presents the long awaited BSG guidelines on IBD. The authors are keen to maintain an interactive dialogue with readers and users of the guidelines and have imbedded links to a discussion forum within the new guidelines. We invite readers to utilise this interesting facility and this will provide an evolving and thorough peer review of the guidelines (see page 571).

Dysbiosis in Crohn’s patients and their relatives A general dysbiosis of the intestinal microbiota has been described in Crohn’s disease (CD) patients but the characterisation of this has remained superﬁcial. There is also no data about the microbiota in relatives of

affected subjects. In this issue of Gut, Joossens et al provide some new insight. Focusing on families with at least three members affected with CD, faecal samples of 68 patients with CD, 84 of their unaffected relatives and 55 matched controls were subjected to community ﬁngerprinting of the predominant microbiota using denaturing gradient gel electrophoresis (DGGE). They were able to describe a dysbiosis signature associated with CD, characterised by ﬁve bacterial species, namely Dialister invisus, an uncharacterised species of Clostridium cluster XIVa, Faecalibacterium prausnitzii, Biﬁdobacterium adolescentis and Ruminococcus gnavus. This dysbiosis signature was markedly characteristic for the disease as it was not observed in unaffected relatives despite a common genetic background and shared nutritional habits. Interestingly, the relatives had their own unique signature which was different from healthy controls. This ﬁrst detailed

Sensitivity (per-patient) of radiologists in detection of lesions of different types and sizes at computed 632 tomographic colonoscopy 633 (CTC), according to performance at initial training, and subsequent experience Sensitivity of radiologists in lower half of distribution

Sensitivity of radiologists in upper half of distribution

Variable of interest

Lesion group

%

95% CI

%

95% CI

Training performance: % detection of polyps of any size in training set (median was 49%)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

52 62 68 61* 68

44 50 48 47 47

to to to to to

60 74 84 74 85

58 76 86 80* 87

51 65 71 68 72

to to to to to

65 85 95 89 96

Training performance: % detection of polyps $6 mm in training set (median was 61%)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

51 63 69 60* 68

43 50 49 46 47

to to to to to

59 74 85 74 85

58 75 85 79* 87

52 64 71 67 72

to to to to to

65 84 94 88 96

Case volume in the study (median number was 18)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

55 56 50 55 25*

40 to 69 31 to 78 12 to 88 23 to 83 1 to 81

55 71 81 72 83*

50 63 70 63 71

to to to to to

61 79 90 81 92

Potential number of polyps for detection (median number was 19)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

58 68 56 70 57

43 47 21 46 18

55 70 82 71 82

49 61 70 61 70

to to to to to

61 78 91 80 91

to to to to to

72 85 86 88 90

The radiologists were divided into two subgroups, according to the median of the distribution of the variable of interest, and the pooled sensitivity was found for each subgroup. *p<0.03 (c2 test and Fisher exact test).

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Digest

Natural history of cirrhosis and acute kidney injury (AKI).

6

HCC xenograft models were established by subcutaneous injection of 5.0310 MHCC cells per mouse in three groups of nude mice (n¼5/group). description of CD-associated dysbiosis will break new ground in unravelling the role of bacteria in CD (see page 631).

A real-life assessment of the performance of CT colonography The assessment of the accuracy of computed tomographic colonography (CTC) has mainly come from studies involving expert radiologists at single academic centres. Heresbach et al have performed a more real-to-life study of CT colonography by investigating its performance in 26 academic clinical units by 28 radiologists. All the radiologists in this study attended a short training course and worked through a training set of 52 cases before beginning to read the CTC exams. In this study, 845 average or high-risk patients were assessed by both complete CTC and videocolonoscopy. Using videocolonoscopy as the gold standard, the positive and negative predictive values for polyps >6 mm in size was 67% (59e74) and 92% (90e94), respectively. Surprisingly, the detection rate for polyps $6 mm was not linked to radiologist

case volume or to number of polyps, but was related to the sensitivity achieved by the radiologist in the training set. This study highlights the need for pre-certiﬁcation of radiologists reading CT colonographic studies in order to ensure high quality CT colonography exams (see page 658).

Hepatology Treatment of hepatocellular carcinoma (HCC)dyet another success of nanotechnology Conventional chemotherapy is rather ineffective for the treatment of HCC. This interesting study investigates a novel approach in vitro and in an animal model of HCC. Ceramide is a very hydrophobic sphingolipid with remarkable proapoptotic properties. The authors constructed ceramide containing nanoliposomes suitable for intravenous application. Using this innovative tool they achieved a marked reduction of tumour size (see ﬁgure) and tumour vascularisation, possibly mediated by interference with akt signalling. Future clinical evalu-

ation of this new therapeutic strategy is to be expected (see page 710).

Renal dysfunction in cirrhosisda never ending story? Renal dysfunction is a major complication in patients with cirrhosis of the liver. The most severe forms of functional acute and chronic renal failure in cirrhosis have been deﬁned as hepatorenal syndrome (HRS) type 1 and type 2, respectively (Gut 2007;56:1310e18). In recent years interest has focused on the functional nature of HRS (see ﬁgure) and an effective treatment with vasopressors in combination with albumin has been established. However, there is increasing recognition, gained for example by screening patients for HRS, that renal dysfunction in cirrhosis can also be due to organic renal diseases and structural changes of the kidney. As yet there is no classiﬁcation system as used in nephrology for such patients. An international working group of outstanding nephrologists and hepatologists has now elaborated a classiﬁcation system for acute kidney injury, chronic kidney disease and acute on chronic kidney disease in cirrhosis. The new classiﬁcation should stimulate future studies of these subtypes of renal dysfunction and may help to elaborate treatment strategies (see page 702).

Gut May 2011 Vol 60 No 5

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Commentary

The benefits of diagnostic ERCP in autoimmune pancreatitis Markus M Lerch, Julia Mayerle Autoimmune pancreatitis (AIP) is a puzzling disease. Although its cardinal features were ﬁrst reported in Europe,1 it seemed, for a long time an Asian phenomenon which many Western clinicians felt they could largely ignore.2 3 As more case series were reported from Europe and America it became increasingly clear that AIP affects patients from many ethnic backgrounds. AIP shares features with two other disorders of the pancreas from which a distinction is critical and determines appropriate treatment. The ﬁrst is chronic pancreatitis of either the environmentally induced (alcohol, tobacco smoke) or the hereditary variety.4e6 The most important difference in terms of treatment, prognosis and, to a lesser degree, diagnosis is that AIP rapidly responds to the administration of steroids7 and other varieties of pancreatitis do not. Reports from Japan suggesting that AIP can be distinguished from chronic pancreatitis by serological markers alone, most prominently serum IgG4 levels,8 were greeted with great enthusiasm. Unfortunately, clinical chemistry was quickly found to be much less reliable in Caucasian patients.9 The reason behind this difference lies in two subtypes of AIP with different prevalence in Europe and Asia. The ﬁrst subtype, and by far the most common in Asia, has recently been termed lymphoplasmacytic sclerosing pancreatitis (LPSP or type I AIP) according to its histological features.10 It is commonly associated with immunological changes such as elevated IgG4 serum levels or various autoantibodies of lesser diagnostic value. A second disease variety named idiopathic duct-centric pancreatitis (IDCP or type II AIP), that barely exists in Japan, accounts for a signiﬁcant number of Caucasian patients, displays often none of the immunological changes and is characterised histologically by granulocytic Department of Medicine A, Greifswald University Hospital, Greifswald, Germany Correspondence to Dr Markus M Lerch, Department of Medicine A, Greifswald University Hospital, Friedrich-Loeffler-Str. 23A, 17475 Greifswald, Germany; [email protected] Gut May 2011 Vol 60 No 5

epithelial lesions.11 This limits diagnostic options for type II AIP to either histology from resection specimens, core biopsies obtained by endoscopic ultrasound, or cross sectional diagnostic imaging. The features of diagnostic imaging, however, are not nearly as well deﬁned or universally established as one would wish. This brings us to the second disease AIP needs to be distinguished from which is pancreatic cancer. The characteristic imaging appearance of AIP has been reported to include diffuse swelling of the entire organ (the ‘sausage shaped’ pancreas) and a diffuse narrowing of the pancreatic duct, often combined with similar changes in the bile ducts when systemic IgG4-associated autoimmune disease is present.12 The latter appearance can easily mimic the ‘double duct’ sign of pancreatic cancer.13 When inﬂammatory inﬁltrates manifest themselves as focal pancreatic enlargement a distinction between cancer and AIP becomes all the more difﬁcult. A focal enlargement was present in no less than 40% of AIP patients in the trial discussed below. This difﬁculty in distinguishing between pancreatic cancer and AIPdparticularly of the IDCP, type II or European varietydremains the principal reason why many patients with AIP still undergo unnecessary pancreatic resection, only to learn after histological examination that surgery was neither required nor will it cure their disease. Until speciﬁc immunological tests become available for type II AIP antigens such as UBR214 or trypsin,15 16 or until better imaging modalities will permit a safer distinction between pancreatitis and cancer the surgical resection rate in patients with AIP will probably not decline. With the latter solution in mind, Sugumar and co-workers17 started their study published in this issue of Gut (see page 666). They assembled a panel of experts from Asia, Europe and the USA and made them read endoscopic retrograde pancreatography (ERP) ﬁlms of conﬁrmed cases of AIP (n¼20), chronic pancreatitis (n¼10) and pancreatic cancer (n¼10). The ﬁrst phase of the study resulted in a rather poor sensitivity for ductal changes, an acceptable speciﬁcity of

92%, and a dismal inter-observer agreement between experts. Phase I of the trial could, however, identify four fairly reliable features for differentiating cancer from AIP. The ﬁrst was the length of the stricture in the pancreatic duct. Long strictures involving more than one third of the duct length, strictures that did not result in an upstream dilatation of the duct, or strictures from which side branches arose were more likely to be caused by AIP than by cancer. The fourth characteristic was the presence of multiple strictures in the duct. The authors went on to generate a teaching module for ERP reading, trained the same 21 experts and asked them to reevaluate the same ERP ﬁlms 3 months later in a blinded fashion. Apparently the teaching model had improved sensitivity to 71% without compromising speciﬁcity but the inter-observer agreement remained poor (0.40). A number of important points can be concluded not only from the study results but also from its limitations: 1. We do not learn from the study what the bile ducts of included patients looked like. Most patients in the study must have undergone endoscopic retrograde cholangio-pancreatography (ERCP) because 71% had presented with painless jaundice. Viewing ERC ﬁlms side by side with corresponding ERP ﬁlms may have increased the sensitivity, speciﬁcity and interobserver agreement greatly. It does not reﬂect clinical practice to limit the evaluating physician to seeing only the latter. 2. The study shows that Asian physicians have much more expertise in diagnosing AIP than European or US-American pancreas experts. One explanation is that the incidence of AIP in Japan and Korea is between ten- and 100-fold higher. Another reason is that type I AIP, the variety most often diagnosed using IgG4 levels and without need for histology, accounts for >95% of patients in Asia. Asian physicians therefore have a diagnostic test at hand to immediately conﬁrm or refute their ERCP interpretation. In the present study 30% of patients were included without diagnostic IgG4 levels. Unfortunately, we do not learn from the paper how many of these were recruited before IgG4 became the marker of choice of type I AIP or whether they all represented type II (IDCP) cases. The current paradox is the following. The Japanese consensus guidelines have made ERCP a mandatory diagnostic 565

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Commentary

criterion although an alternative test (serum IgG4) would be diagnostic in the majority of Japanese patients.18 Conversely, the HISORt criteria from the Mayo Clinic,19 derived from experience with mostly Caucasian patients, do not include ERCP as a mandatory diagnostic test although Caucasian patients with AIP often present without diagnostic lab tests and are much more likely to proﬁt from diagnostic ERCP because core biopsy is the only alternative test for conﬁrming the diagnosis (short of obtaining resection specimens) in the prevalent type II AIP. Not surprisingly, a study with mostly Western patients relied for differentiation between cancer and AIP on a protocol that included ERCP20 whereas a Japanese study addressing the same issue made do without it.21 What remains to be addressed in future studies is how well the ERP criteria of Sugumar et al work speciﬁcally in type II autoimmune pancreatitis (IDCP) because this is the subgroup which potentially beneﬁts the most from diagnostic ERCP and in which ERCP may prevent most of the unnecessary pancreatic resections. The studies also need to clarify whether concerns about post-ERCP pancreatitis in patients with suspected AIP are unfounded. The good news of the study is that nonAsian pancreas experts can greatly improve their skills in reading ERP-ﬁlms if given an appropriate teaching module. Moreover, ER(C)P was conﬁrmed as a powerful tool to diagnose AIP and to distinguish it from either chronic pancreatitis or pancreatic cancer. The study supports the Japanese consensus which strongly recommends ERCP for the diagnosis of AIP18 and argues against the Mayo view19 which does not make it a requirement. If the ductal changes reported here can be conﬁrmed in their speciﬁcity for type II autoimmune pancreatitis, as well as the superiority of

566

ERCP over MRCP in detecting them, then AIP will side with primary sclerosing cholangitis as a condition in which diagnostic ERCP cannot yet be replaced by less invasive diagnostic test. Whether ERCP can serve in a similar role for monitoring treatment response in AIP will also have to be addressed in future trials. Funding The author’s own work is supported by the Alfried-Krupp-von-Bohlen-und-Hallbach-Foundation (Graduate Schools Tumour Biology and Free Radical Biology), the Deutsche Krebshilfe/ Dr. Mildred-Scheel-Stiftung (109102), the Deutsche Forschungsgemeinschaft (DFG GRK840-E3/E4, MA 4115/1-2/3, NI 1297/1-1), the Federal Ministry of Education and Research (BMBF GANI-MED 03152061A and BMBF 0314107) and the European Union (EU-FP-7: EPC-TM and EU-FP7-REGPOT-2010-1).

8.

9.

10.

11.

12.

Competing interests None declared. Patient consent Obtained.

13.

Provenance and peer review Commissioned; externally peer reviewed. Published Online First 11 February 2011

14.

Gut 2011;60:565e566. doi:10.1136/gut.2010.232157 15.

REFERENCES 1.

2.

3.

4.

5.

6.

7.

Sarles H, Sarles JC, Muratore R, et al. Chronic inflammatory sclerosis of the pancreas e an autonomous pancreatic disease? Am J Dig Dis 1961;1:688e98. Yoshida K, Toki F, Takeuchi T, et al. Chronic pancreatitis caused by an autoimmune abnormality. Proposal of the concept of autoimmune pancreatitis. Dig Dis Sci 1995;40:1561e8. Varadarajulu S, Cotton PB. Autoimmune pancreatitis: is it relevant in the west? Gastroenterology 2003;125:1557. Pickartz T, Mayerle J, Lerch MM. Autoimmune pancreatitis. Nat Clin Pract Gastroenterol Hepatol 2007;4:314e23. Keim V, Bauer N, Teich N, et al. Clinical characterization of patients with hereditary pancreatitis and mutations in the cationic trypsinogen gene. Am J Med 2001;111:622e6. Gress TM, Mu¨ller-Pillasch F, Lerch MM, et al. Balance of expression of genes coding for extracellular matrix proteins and extracellular matrix degrading proteases in chronic pancreatitis. Z Gastroenterol 1994;32:221e5. Moon SH, Kim MH, Park DH, et al. Is a 2-week steroid trial after initial negative investigation for

16.

17.

18.

19.

20.

21.

malignancy useful in differentiating autoimmune pancreatitis from pancreatic cancer? A prospective outcome study. Gut 2008;57:1704e12. Hamano H, Kawa S, Horiuchi A, et al. High serum IgG4 concentrations in patients with sclerosing pancreatitis. N Engl J Med 2001;344:732e8. Aparisi L, Farre A, Gomez-Cambronero L, et al. Antibodies to carbonic anhydrase and IgG4 levels in idiopathic chronic pancreatitis: relevance for diagnosis of autoimmune pancreatitis. Gut 2005;54:703e9. Chari ST, Kloeppel G, Zhang L, et al; Autoimmune Pancreatitis International Cooperative Study Group (APICS). Histopathologic and clinical subtypes of autoimmune pancreatitis: the Honolulu consensus document. Pancreas 2010;39:549e54. Klo¨ppel G, Detlefsen S, Chari ST, et al. Autoimmune pancreatitis: the clinicopathological characteristics of the subtype with granulocytic epithelial lesions. J Gastroenterol 2010;45:787e93. Ghazale A, Chari ST, Zhang L, et al. Immunoglobulin G4-associated cholangitis: clinical profile and response to therapy. Gastroenterology 2008;134:706e15. Menges M, Lerch MM, Zeitz M. The double duct sign in patients with malignant and benign pancreatic lesions. Gastrointest Endosc 2000;52:74e7. Frulloni L, Lunardi C, Simone R, et al. Identification of a novel antibody associated with autoimmune pancreatitis. N Engl J Med 2009;361:2135e42. Halangk W, Kru¨ger B, Ruthenbu¨rger M, et al. Trypsin activity is not involved in premature, intrapancreatic trypsinogen activation. Am J Physiol Gastrointest Liver Physiol 2002;282:G367e74. Lo¨hr JM, Faissner R, Koczan D, et al. Autoantibodies against the exocrine pancreas in autoimmune pancreatitis: gene and protein expression profiling and immunoassays identify pancreatic enzymes as a major target of the inflammatory process. Am J Gastroenterol 2010;105:2060e71. Sugumar A, Levy MJ, Kamisawa T, et al. Endoscopic retrograde pancreatography criteria to diagnose autoimmune pancreatitis: an international multicenter study. Gut 2011;60:666e70. Kamisawa T, Okazaki K, Kawa S, et al. Japanese consensus guidelines for management of autoimmune pancreatitis: III. Treatment and prognosis of AIP. J Gastroenterol 2010;45:471e7. Chari ST. Diagnosis of autoimmune pancreatitis using its five cardinal features: introducing the Mayo Clinic’s HISORt criteria. J Gastroenterol 2007;42 (Suppl 18):39e41. Chari ST, Takahashi N, Levy MJ, et al. A diagnostic strategy to distinguish autoimmune pancreatitis from pancreatic cancer. Clin Gastroenterol Hepatol 2009;7:1097e103. Kamisawa T, Imai M, Yui Chen P, et al. Strategy for differentiating autoimmune pancreatitis from pancreatic cancer. Pancreas 2008;37:e62e7.

Gut May 2011 Vol 60 No 5

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Commentary

Long-term proton pump inhibitor administration, H pylori and gastric cancer: lessons from the gerbil James G Fox,1 Ernst J Kuipers2 The association between chronic active gastritis and pre-neoplastic conditions as well as invasive cancer of the stomach was established several decades ago. The risk of progression depended on the severity and distribution of gastritis, with cancer, in particular, occurring in subjects with pan-gastritis. Subsequently, Helicobacter pylori was recognised as the primary cause of chronic active gastritis, and it was demonstrated that the pattern of gastritis corresponded with the colonisation pattern of H pylori. We and others then showed both in animals and in humans that this pattern of colonisation and associated gastritis primarily depended on the level of acid output.1 2 Although this hypothesis was widely accepted, it led to intense debate when dealing with the safety of long-term treatment with profound acid suppressors. An elegant, long-awaited study from Japan published in this issue of Gut (see page 624) provides compelling evidence that the pattern of H pylori colonisation depends on acid output and that this inﬂuences the long-term progression to neoplasia.3 The current study was based on experiments in gerbils, one of the well-established animal models for the study gastric disease induced by Helicobacter spp.4 Gerbils, used sparingly for biomedical research, were ﬁrst reported as a model for experimental H pylori infection in 1991.5 Interestingly, the gerbil has been shown to have particular relevant features that can be used to address whether H pylori can induce gastric cancer. Japanese investigators

1

Division of Comparative Medicine and Department of Biological Engineering, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA; 2 Departments of Internal Medicine, and Gastroenterology & Hepatology, Erasmus Medical Center, Rotterdam, The Netherlands Correspondence to Dr James G Fox, Division of Comparative Medicine, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Bldg 16-825, Cambridge, MA 02139, USA; [email protected] Gut May 2011 Vol 60 No 5

have noted intestinal metaplasia, atrophy and gastric ulcers in gerbils experimentally infected with H pylori.6 7 Following these ﬁndings, others noted that gerbils infected with H pylori from periods ranging from 15 to 18 months develop gastric adenocarcinoma.8 9 The gastric cancers were clearly documented histologically. Vascular invasion and metastases were not observed in either study. The histological progression of H pyloriassociated disease in the gerbil closely resembles that observed in humans, including the early appearance of intestinal metaplasia, well-differentiated histological patterns of gastric malignancy, and antral location of the gastric cancers. The development of cancer in the gerbil is also preceded by invagination of atypical glands (cystica profunda) into the submucosa. The association with gastric ulcers in this model continues to be of interest, given clinical studies in human patients indicating a link between gastric ulcer disease and gastric cancer.10 Although most of the tumours in the H pylori gerbil model originate in the pyloric region of the stomach, signiﬁcant changes in the oxyntic mucosa consistent with chronic atrophic gastritis are also observed. Glandular tissues in the gastric body and fundus are replaced by atrophy and hyperplastic epithelium of the pseudopyloric type. A similar type of gastric atrophy (loss of oxyntic glands and neck cell hyperplasia) has also been reported in Helicobacter felis or H pylori-infected C57BL mice.11 12 The gerbil appears to be uniquely susceptible to H pylori-induced gastric neoplasia, and further characterisation of the gerbil model has provided important clues on gastric cancer progression and hostebacteria interaction. The unusual susceptibility of this animal species to gastric cancer using fairly standard H pylori strains underscores once again the over-riding importance of host factors in determining the outcome from gastric

Helicobacter spp. infection.13 Further, a possible role for altered gastrin physiology in the pathogenesis of gastric cancer has been raised by several recent studies. In Mongolian gerbils, H pylori leads to marked elevation of serum gastrin levels which coincide temporally with increases in gastric mucosal proliferation rates.14 This is consistent with experiments in inbred (FVB/N) mice that were rendered moderately hypergastrinaemic through an insulinegastrin (INS-GAS) transgene. In this model there is increased mucosal proliferation and progressive atrophy, and gastric carcinomas develop in 20-monthold INS/GAS mice.15 Infection of these hypergastrinaemic mice with H felis and H pylori accelerates tumorigenesis, with the majority (85%) of infected male mice developing gastric cancer within 8 months after colonisation. We recently reported that this chronic, progressive process is enhanced by co-colonisation with enteric ﬂora.16 Using this well-described and reproducible gerbil model, Japanese investigators have conducted a very important study, examining the potential effect of proton pump inhibitor (PPI) therapy on progression of H pylori-associated disease.3 The male gerbils were divided into four groups: H pylori (ATCC43504) positives and negatives, with and without administration of a PPI (omeprazole 100 mg/kg body weight/day). At the end point of the 6-month experiment, the authors provided detailed analysis of gastric pathology including morphometric severity of parietal cell loss. The authors made a concerted effort to address confusion in the literature with respect to what constitutes a ‘gastric adenocarcinoma’ in a Mongolian gerbil model and how to differentiate an ‘invasive carcinoma’ from ‘heterotopic glands’ that are frequently encountered in these models during the early course of the disease. The ability to differentiate glandular herniation into the submucosa from the true invasion is an important consideration in evaluating rodent models of gastric cancer induced by Helicobacter spp. Indeed with these histopathological criteria in mind the results were remarkable and provide considerable support for previously published literature in humans. H pylori-negative male gerbils did not develop gastritis, metaplasia, or cancer irrespective of PPI treatment. H pylori-positive gerbils all developed gastritis, and most also developed metaplasia during the later stages of disease. Treatment with omeprazole in H pyloripositive animals accelerated progression to atrophy and enhanced hypergastrinaemia, 567

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Commentary

resulting in a signiﬁcantly increased incidence of gastric cancer.3 The authors emphasise that hypergastrinaemia might be promoting the development of H pylori gastric cancer. Given gender susceptibility to H pylori-associated gastric cancer in humans and rodent models, additional studies using H pylori female gerbils on chronic PPI therapy should provide additional interesting insights. Overall, the ﬁndings in the current paper support previous ﬁndings by us and others showing the increased risk of gastric atrophy in H pylori-infected patients on PPI therapy.2 17e19 In humans, it remains to be shown whether chronic PPI therapy in H pylori-infected individuals also increases the further progression to metaplasia, dysplasia and neoplasia. These studies will require longer-term follow-up of this subset of patients. In the meantime, the current ﬁndings in this paper support the argument regarding the importance of considering H pylori eradication in patients on long-term PPI treatment to cure gastritis, lower gastrin levels, and prevent progression to atrophy, dysplasia and gastric cancer. This recommendation is consistent with European guidelines on H pylori treatment in humans.20 Competing interests None. Provenance and peer review Commissioned; not externally peer reviewed.

Published Online First 17 February 2011

10.

Gut 2011;60:567e568. doi:10.1136/gut.2010.229286 11.

REFERENCES 1.

2.

3.

4. 5.

6.

7.

8.

9.

Danon SJ, O’Rourke JL, Moss ND, et al. The importance of local acid production in the distribution of Helicobacter felis in the mouse stomach. Gastroenterology 1995;108:1386e95. Kuipers EJ, Uyterlinde AM, Pena AS, et al. Increase of Helicobacter pylori-associated corpus gastritis during acid suppressive therapy: implications for long-term safety. Am J Gastroenterol 1995;90:1401e6. Hagiwara T, Mukaisho K, Nakayama T, et al. Long-term proton pump inhibitor administration worsens atrophic corpus gastritis and promotes adenocarcinoma development in Mongolian gerbils infected with Helicobacter pylori. Gut 2011;60:624e30. Fox JG, Wang TC. Inflammation, atrophy, and gastric cancer. J Clin Invest 2007;117:60e9. Yokota K, Kurebayashi Y, Takayama Y, et al. Colonization of Helicobacter pylori in the gastric mucosa of Mongolian gerbils. Microbiol Immunol 1991;35:475e80. Hirayama F, Takagi S, Kusuhara H, et al. Induction of gastric ulcer and intestinal metaplasia in mongolian gerbils infected with Helicobacter pylori. J Gastroenterol 1996;31:755e7. Honda S, Fujioka T, Tokieda M, et al. Gastric ulcer, atrophic gastritis, and intestinal metaplasia caused by Helicobacter pylori infection in Mongolian gerbils. Scand J Gastroenterol 1998;33:454e60. Honda S, Fujioka T, Tokieda M, et al. Development of Helicobacter pylori-induced gastric carcinoma in Mongolian gerbils. Cancer Res 1998;58:4255e9. Watanabe T, Tada M, Nagai H, et al. Helicobacter pylori infection induces gastric cancer in mongolian gerbils. Gastroenterology 1998;115:642e8.

12.

13.

14.

15.

16.

17.

18.

19. 20.

Hansson LE, Nyren O, Hsing AW, et al. The risk of stomach cancer in patients with gastric or duodenal ulcer disease. N Engl J Med 1996;335:242e9. Wang TC, Goldenring JR, Dangler C, et al. Mice lacking secretory phospholipase A2 show altered apoptosis and differentiation with Helicobacter felis infection. Gastroenterology 1998;114:675e89. Fox JG, Dangler CA, Taylor NS, et al. High-salt diet induces gastric epithelial hyperplasia and parietal cell loss, and enhances Helicobacter pylori colonization in C57BL/6 mice. Cancer Res 1999;59:4823e8. Wang TC, Fox JG. Helicobacter pylori and gastric cancer: Koch’s postulates fulfilled? Gastroenterology 1998;115:780e3. Peek RM Jr, Wirth HP, Moss SF, et al. Helicobacter pylori alters gastric epithelial cell cycle events and gastrin secretion in Mongolian gerbils. Gastroenterology 2000;118:48e59. Wang TC, Dangler CA, Chen D, et al. Synergistic interaction between hypergastrinemia and Helicobacter infection in a mouse model of gastric cancer. Gastroenterology 2000;118:36e47. Lofgren JL, Whary MT, Ge Z, et al. Lack of commensal flora in Helicobacter pylori-infected INSGAS mice reduces gastritis and delays intraepithelial neoplasia. Gastroenterology 2011;140:210e20. Kuipers EJ, Lundell L, Klinkenberg-Knol EC, et al. Atrophic gastritis and Helicobacter pylori infection in patients with reflux esophagitis treated with omeprazole or fundoplication. N Engl J Med 1996;334:1018e22. Lundell L, Havu N, Miettinen P, et al. Changes of gastric mucosal architecture during long-term omeprazole therapy: results of a randomized clinical trial. Aliment Pharmacol Ther 2006;23:639e47. Kuipers EJ. Proton pump inhibitors and gastric neoplasia. Gut 2006;55:1217e21. Malfertheiner P, Megraud F, O’Morain C, et al. Current concepts in the management of Helicobacter pylori infection: the Maastricht III consensus report. Gut 2007;56:772e81.

Gut table of contents alerts Sign up for our electronic table of contents alert and you will never miss new issues of Gut when published online. Stay ahead and up to date by visiting gut.bmj.com.

568

Gut May 2011 Vol 60 No 5

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Commentary

Breaking down haem attenuates acute pancreatitis: a new treatment option? Julia Mayerle, Matthias Sendler, Markus M Lerch The premature intracellular activation of digestive proteases can result in autodigestion of the pancreas and has long been regarded as the initial and a rate-limiting step in the development of acute pancreatitis.1 Recent evidence suggests, however, that inﬂammatory cells inﬁltrating the pancreasdrather than the extent of initial protease activationddetermines the ultimate severity of acute pancreatitis.2 3 All attempts to target the inﬂammatory response of acute pancreatitis therapeutically have unfortunately been unsuccessful. One reason for this failure may lie in the type of inﬂammatory cells that were targeted.4 Macrophages orchestrate both the initiation and the resolution of inﬂammation and would therefore be an interesting therapeutic target for modulating the inﬂammatory response in acute pancreatitis. Selective depletion of the proinﬂammatory M1 macrophages has been found to reduce the severity in acute pancreatitis.5 Nature has provided a potent protective strategy to counter an overwhelming inﬂammatory response. Haemoxygenase-1 (HO-1) is an inducible isoform of the initial and rate-controlling enzyme used in the degradation of haem into iron, carbon monoxide and biliverdin. The latter is subsequently converted to bilirubin.6 HO-1 is expressed in various inﬂammatory cells including macrophages and has recently gained considerable attention for exerting anti-inﬂammatory, anti-apoptotic, angiogenic and cytoprotective actions. Haem that is either released from resolved erythrocytes or therapeutically applied as haemin (or in its water soluble form panhaematin) induces HO-1 expression in inﬂammatory cells. HO-1 will, in turn, break down the haem and trigger different cellular events: (1) the haem breakdown product carbon monoxide is a potent Department of Medicine A, Ernst-Moritz-Arndt-University Greifswald, Greifswald, Germany Correspondence to Dr Julia Mayerle, Department of Medicine A, Ernst-Moritz-Arndt-University, Greifswald, Friedrich-Loeffler-Str 23a, Greifswald 17475, Germany; [email protected] Gut May 2011 Vol 60 No 5

vasodilator and can resolve the microcirculatory disturbances associated with acute pancreatitis and promote angiogenesis during the recovery from necrosis. It will also mediate anti-apoptotic effects via the activation of the PERK-Nrf2-HO-1 axis and PERK-eIF2a-ATF4 pathways resulting in the downregulation of CHOP (DNA damage-inducible transcript 3); (2) haem is also broken down into bilirubin which can act as a potent antioxidant; and (3) during haem degradation, iron is released which is then sequestered into the acute phase protein ferritin.7 HO-1 is induced in macrophages as early as 2 h after intraperitoneal or intravenous haemin or panhaematin treatment. Macrophages then home to the inﬂamed pancreas and start the resolution phase of pancreatitis. Macrophages with induced HO-1 activity respond to lipopolysaccharide (LPS) stimulation with decreased secretion of proinﬂammatory cytokines interleukin (IL)-6, tumour necrosis factor a, IL-1b and macrophage inﬂammatory protein 1b but increased IL-10 secretion. The latter results in an autoregulatory loop in which IL-10 again induces HO-1 and can thereby protect mice from LPS-induced septic shock.6 IL-10-mediated HO-1 activity may also add to the beneﬁcial role of IL-10 treatment in acute pancreatitis.8 In this manner, HO-1 contributes to the shift from M1 (proinﬂammatory) to M2 (anti-inﬂammatory, proﬁbrogenic) macrophages.9 HO-1 upregulation reduces the rolling, sticking and transmigration of inﬂammatory cells from the vascular compartment.6 This immunoregulatory role of HO-1 has already been implicated in a variety of diseases ranging from type 1 diabetes and allergic asthma to different infectious diseases such as malaria, listeria, salmonella and pneumococcal infections. Ischaemiareperfusion damage and acute or chronic rejection after organ transplantation are also affected by HO-1 activity.6 In 2005 Omary et al showed that prophylactic haemin treatment reduces the mortality in choline-deﬁcient-dietinduced (CDD) pancreatitis in mice from 45% to 0%.10 HO-1 upregulation in

macrophages occurred as early as 2 h after haemin treatment and was shown to be responsible for the beneﬁcial treatment effect. Translation to the human situation was accomplished late last year when the same group demonstrated that HO-1 is upregulated in peripheral blood mononuclear cells during the course of pancreatitis and that monocytes from patients with acute pancreatitis are primed for HO-1 induction by panhaematin treatment.11 The ﬁnal evidence for a beneﬁcial action of panhaematin in the treatment of experimental pancreatitis is reported in this issue of Gut (see page 671).12 In this study, Habtezion and co-workers used two different models of acute pancreatitis (CDD, L-arginine) and treated animals with panhaematin up to 24 h after disease onset. In this late treatment window of pancreatitis, which corresponds to a frequent clinical situation, the authors achieved a signiﬁcant reduction in mortality. They excluded a direct effect of panhaematin or of macrophages with upregulated HO-1 on pancreatic acinar cells and concluded that the beneﬁcial effect is the result of reduced inﬁltration of inﬂammatory cells. This was supported by the results of CD45 staining of the pancreas and reduced levels of the neutrophil chemoattractant CXCL1. What remained unresolved in this is study is whether the antioxidant effect of bilirubin13 or a vascular effect of carbon monoxide also added to the beneﬁcial effect of panhaematin treatment. Antioxidant treatment has been suggested to prevent organ failure in severe acute pancreatitis,14 but clinical trials have reported conﬂicting results.15 16 The therapeutic potential of antioxidants in clinical acute pancreatitis has therefore remained an open question. Regardless of the underlying mechanism by which the upregulation of HO-1 in macrophages or the therapeutic application of panhaematin exert a beneﬁcial effect on the course of two animal models of severe pancreatitis, Habtezion and co-workers have established haem as a novel treatment option for pancreatitis. Haem, or its soluble pharmaceutical equivalent panhaematin or haemin, is already an FDA approved agent for the treatment of acute intermittent porphyria. Prospective clinical trials for pancreatitis are therefore a realistic possibility in the near future. Whether the presently substantial treatment costs of haemin or its nephrotoxicity will limit the enthusiasm for introducing this interesting pathophysiological discovery 569

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Commentary

of Habtezion and co-workers into clinical practice will remain to be seen. Funding The authors’ own work is supported by the Alfried-Krupp-von-Bohlen-und-Hahlbach-Foundation (Graduate Schools Tumour Biology and Free Radical Biology), the Deutsche Krebshilfe/ Dr. Mildred-Scheel-Stiftung (109102), the Deutsche Forschungsgemeinschaft (DFG GRK840-E3/E4, MA 4115/1-2/3, NI 1297/1-1), the Federal Ministry of Education and Research (BMBF GANI-MED 03152061A and BMBF 0314107) and the European Union (EU-FP-7: EPC-TM and EU-FP7-REGPOT-2010-1).

3.

4.

5.

6.

Competing interests None. Provenance and peer review Commissioned; externally peer reviewed. Published Online First 11 February 2011

7.

8.

Gut 2011;60:569e570. doi:10.1136/gut.2010.230136

REFERENCES 1.

2.

Ruthenburger M, Mayerle J, Lerch MM. Cell biology of pancreatic proteases. Endocrinol Metab Clin North Am 2006;35:313e31. Mayerle J. A novel role for leucocytes in determining the severity of acute pancreatitis. Gut 2009;58:1440e1.

9.

10.

Halangk W, Kruger B, Ruthenburger M, et al. Trypsin activity is not involved in premature, intrapancreatic trypsinogen activation. Am J Physiol Gastrointest Liver Physiol 2002;282:G367e74. Mayerle J, Schnekenburger J, Kruger B, et al. Extracellular cleavage of E-cadherin by leukocyte elastase during acute experimental pancreatitis in rats. Gastroenterology 2005;129:1251e67. Sendler MDA, Weiss FU, Brandt-Nedelev B, et al. Tissue damage and protease activation in acute pancreatitis via TNF alpha secretion. Pancreas 2009;38:1045. Blancou P, Tardif V, Simon T, et al. Immunoregulatory properties of heme oxygenase-1. Methods Mol Biol 2011;677:247e68. Kim HP, Pae HO, Back SH, et al. Heme oxygenase-1 comes back to endoplasmic reticulum. Biochem Biophys Res Commun 2011;404:1e5. Deviere J, Le Moine O, Van Laethem JL, et al. Interleukin 10 reduces the incidence of pancreatitis after therapeutic endoscopic retrograde cholangiopancreatography. Gastroenterology 2001;120:498e505. Shrivastava P, Bhatia M. Essential role of monocytes and macrophages in the progression of acute pancreatitis. World J Gastroenterol 2010;16:3995e4002. Nakamichi I, Habtezion A, Zhong B, et al. Hemin-activated macrophages home to the pancreas

11.

12.

13.

14. 15.

16.

and protect from acute pancreatitis via heme oxygenase-1 induction. J Clin Invest 2005;115:3007e14. Habtezion A, Kwan R, Yang AL, et al. Heme oxygenase-1 is induced in peripheral blood mononuclear cells of patients with acute pancreatitis: a potential therapeutic target. Am J Physiol Gastrointest Liver Physiol 2011;300:G12e20. Habtezion A, Kwan R, Akhtar E, et al. Panhematin provides a therapeutic benefit in experimental pancreatitis. Gut 2011;60:671e79. MacLean PD, Drake EC, Ross L, et al. Bilirubin as an antioxidant in micelles and lipid bilayers: its contribution to the total antioxidant capacity of human blood plasma. Free Radic Biol Med 2007;43:600e9. Johnson CD. Antioxidants in acute pancreatitis. Gut 2007;56:1344e5. Sateesh J, Bhardwaj P, Singh N, et al. Effect of antioxidant therapy on hospital stay and complications in patients with early acute pancreatitis: a randomised controlled trial. Trop Gastroenterol 2009;30:201e6. Siriwardena AK, Mason JM, Balachandra S, et al. Randomised, double blind, placebo controlled trial of intravenous antioxidant (n-acetylcysteine, selenium, vitamin C) therapy in severe acute pancreatitis. Gut 2007;56:1439e44.

Gut online Visit Gut online and listen to the latest podcast, post comments and download any you might have missed. Keep informed and up to date by visiting gut.bmj.com.

570

Gut May 2011 Vol 60 No 5

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines

Guidelines for the management of inflammatory bowel disease in adults Craig Mowat,1 Andrew Cole,2 Al Windsor,3 Tariq Ahmad,4 Ian Arnott,5 Richard Driscoll,6 Sally Mitton,7 Tim Orchard,8 Matt Rutter,9 Lisa Younge,10 Charlie Lees,5 Gwo-tzer Ho,5 Jack Satsangi,5 Stuart Bloom,11 on behalf of the IBD Section of the British Society of Gastroenterology 1

Gastrointestinal Unit, Ninewells Hospital, Dundee, UK 2 Derby Digestive Disease Centre, Royal Derby Hospital, Derby, UK 3 Department of Surgery, University College London Hospitals, UK 4 Department of Gastroenterology, Royal Devon & Exeter Hospital, UK 5 Gastrointestinal Unit, Western General Hospital, Edinburgh, UK 6 Crohn’s and Colitis UK (NACC), St. Albans, UK 7 Department of Paediatrics, St George’s Hospital, London, UK 8 Department of Gastroenterology, St. Mary’s Hospital, London, UK 9 Department Gastroenterology, University Hospital of North Tees, UK 10 Endoscopy Unit, Royal London Hospital, UK 11 Department of Gastroenterology, University College London Hospitals, UK Correspondence to Dr Craig Mowat, Gastrointestinal Unit, Ninewells Hospital & Medical School, Dundee DD1 9SY, UK; [email protected] Revised 22 November 2010 Accepted 7 December 2010

ABSTRACT The management of inflammatory bowel disease represents a key component of clinical practice for members of the British Society of Gastroenterology (BSG). There has been considerable progress in management strategies affecting all aspects of clinical care since the publication of previous BSG guidelines in 2004, necessitating the present revision. Key components of the present document worthy of attention as having been subject to re-assessment, and revision, and having direct impact on practice include: < The data generated by the nationwide audits of inflammatory bowel disease (IBD) management in the UK in 2006, and 2008. < The publication of ‘Quality Care: service standards for the healthcare of people with IBD’ in 2009. < The introduction of the Montreal classification for Crohn’s disease and ulcerative colitis. < The revision of recommendations for the use of immunosuppressive therapy. < The detailed analysis, guidelines and recommendations for the safe and appropriate use of biological therapies in Crohn’s disease and ulcerative colitis. < The reassessment of the role of surgery in disease management, with emphasis on the importance of multi-disciplinary decision-making in complex cases. < The availablity of new data on the role of reconstructive surgery in ulcerative colitis. < The cross-referencing to revised guidelines for colonoscopic surveillance, for the management of metabolic bone disease, and for the care of children with inflammatory bowel disease. < Use of the BSG discussion forum available on the BSG website to enable ongoing feedback on the published document http://www.bsg.org.uk/forum (accessed Oct 2010). The present document is intended primarily for the use of clinicians in the United Kingdom, and serves to replace the previous BSG guidelines in IBD, while complementing recent consensus statements published by the European Crohn’s and Colitis Organisation (ECCO) https://www.ecco-ibd.eu/index.php (accessed Oct 2010).

1.0 INTRODUCTION These guidelines have been commissioned by the Clinical Services and Standards Committee of the British Society of Gastroenterology (BSG) for clinicians and allied professionals caring for patients Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

with inﬂammatory bowel disease in the United Kingdom. The authors of these guidelines were members of the BSG IBD Committee at the time. This committee is elected by fellow members of the IBD section of the Society. They replace the guidelines published in 2004 by Carter et al.1 They have been written with close reference to the recent European evidence-based consensus documents on Crohn’s disease2e4 and ulcerative colitis5e7 produced by the European Crohn’s and Colitis Organisation (ECCO) (https://www.ecco-ibd.eu/ index.php (accessed Oct 2010)). Although these consensus documents provide a comprehensive and authoritative review of the evidence base underlying deﬁnitions, diagnosis and current management we believe there are still compelling reasons to issue a set of up-to-date guidelines for UK practice: 1. These diseases are complex and recent UK-wide audits have demonstrated wide variation in clinical practice.8 9 2. There may be important differences between guidelines and consensus. Clinical practice guidelines are ‘systematically developed statements to assist practitioner and patient decisions about appropriate healthcare for speciﬁc clinical circumstances’. Their speciﬁc purpose is to make explicit recommendations with an intent to inﬂuence what clinicans do.10 Recommendations for practice with particular reference to one country may not always be identical to consensus statements (eg, use of maintenance anti-tumour necrosis factor (TNF) antibodies in Crohn’s disease). 3. The recent publication of a set of UK Service Standards for the healthcare of people who have IBD is very relevant to guidelines for UK practice. http://www.ibdstandards.org.uk (accessed Oct 2010).11 4. UK practice is inﬂuenced by guidance from the National Institute for Clinical Excellence (NICE) http://www.nice.org.uk/ (accessed Oct 2010). 5. In some areas, the approach of UK physicians differs from the European consensus, and indeed from North American practice, and the present document provides an appropriate and necessary addition to the current literature. This is particular necessary at the present time, when new therapies are being introduced in IBD, and previously accepted management paradigms are being extensively revised. 6. Publication of these guidelines will be supported by the establishment of a discussion forum on the 571

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines BSG website to enable ongoing feedback on the published document http://www.bsg.org.uk/forum (accessed Oct 2010).

1.1 Guideline development The guidelines were drafted shortly after the ECCO consensus was published in the knowledge of the extremely rigorous nature and literature review accompanying that process; throughout the current document, reference is made to the ECCO consensus statements. Each author responsible for drafting sections carried out a further literature review and the draft and accompanying evidence base was extensively discussed in committee. The ECCO recommendations were formally compared to the 2004 BSG guidance at the outset, in order to identify areas of disparity in opinion and content. The important issue of assessing guideline quality has been addressed by use of the AGREE tool (see section 1.2). We feel readers will want to assess the recommendations as they would for other guidelines, including a judgement based on accompanying levels of evidence, which are included throughout the text. A preliminary document was drafted by contributing authors (coordinated by CM). Recommendations were submitted by contributing authors and voted on before being incorporated into the guidelines. Draft guidelines were submitted for review by the BSG Clinical Services and Standards Committee before being submitted for further review by BSG council members and simultaneously by external reviewers chosen by the editor of Gut. The format of the ﬁrst edition of the guidelines has largely been retained with modiﬁcations to emphasise aspects of service delivery and patient expectations relevant to UK practice. Sections have been added on UK standards of care and principles of nutrition. Drug therapies have been given a separate section rather than being included in disease management and guidelines have been extensively re-written to take account of developments in use of immunosuppressive and biological treatments.

1.2 Assessing the quality of guidelines: The AGREE instrument There have been attempts to develop external validation for clinical practice guidelines; the best characterised of these is the AGREE instrument developed by the AGREE collaboration (http://www.agreecollaboration.org/intro/ (accessed Oct 2010)). The audit department of the Royal College of Physicians has adopted this tool. It identiﬁes six criteria of quality which are addressed here.

1. Scope and purpose The guidelines are intended for use by clinicians and other healthcare professionals in managing patients with ulcerative colitis and Crohn’s disease in the light of recent guidance published by NICE and the development of the IBD standards of care. They are primarily aimed at management of adult patients: there are separate guidelines published for the care of children with IBD (http://bspghan.org.uk/IBDGuidelines (accessed Oct 2010)). They contain reference to speciﬁc issues relating to management of adolescents with IBD.

2. Guideline development group membership and stakeholder involvement Membership of the group is detailed at the end of the document: it includes medical and surgical gastroenterologists, IBD specialist nurses, members of the British Dietetic Association, 572

and patient representative groups. The section on imaging has been approved by the British Society of Gastrointestinal and Abdominal Radiology committee.

3. Rigour of development The published literature has been searched using Pubmed, Medline and the Cochrane database. The guidelines rely considerably on consensus statements published by the European Crohn’s and Colitis Organisation (ECCO).2e7 In order to harmonise management guidelines with the ECCO consensus statements, we have adopted a similar style of graded recommendations (graded AeD), determined by the level of supporting evidence (graded level 1e5) as described by the Oxford Centre For Evidence Based Medicine (table 1). Areas of disagreement about the recommendation grade were subjected to discussion and if necessary voting by members of the guidelines group. Where possible, the health beneﬁts, side effects and risks of recommendations have been discussed. The guidelines have been peer reviewed according to the editorial policy of Gut.

4. Clarity and presentation Recommendations are intended to be speciﬁc to particular situations and patient groups; where necessary, different options are listed. Key recommendations are linked to discussion threads on a discussion forum hosted on the BSG website.

5. Applicability Where necessary, we have discussed organisational changes that may be needed in order to apply recommendations. We have attempted to identify key criteria for monitoring and audit purposes.

6. Editorial independence and conflict of interest Guideline group members have declared any conﬂicts of interest.

1.3 Scheduled review of guidelines The content and evidence base for these guidelines should be revised within 4 years of publication, to take account of new evidence. We anticipate the guidelines will continue to evolve through evidence gathered by regular national audit of IBD standards and services. Guidelines, by their nature, will become outdated as new evidence is published. With this in mind (and also to provide user feedback) links to the BSG discussion forum relating to speciﬁc sections of these guidelines are included in this document. These forums are accessible to any member of the BSG who are encouraged to contribute citing the appropriate new evidence. In line with the agree tool the BSG forum will also provide some user feedback on the guidelines. The links will be at http://www.bsg.org.uk/forum/ and the links for speciﬁc sections are also imbedded within this document.

2.0 SERVICE DELIVERY 2.1 Impact of inflammatory bowel disease on patients and society With a reported prevalence of 400 per 100 00012 there are approximately 240 000 patients with IBD in the UK (ulcerative colitis: 243/100 000¼146 000 people in UK population of 60 million; Crohn’s disease: 145/100 000¼87 000 people in UK). The incidence of Crohn’s disease in the UK increased markedly between the 1950s and the 1980s.13 14 Since the 1980s the incidence of Crohn’s disease has continued to increase in the UK at a rather slower rate.15 16 Most patients are referred to hospital clinics for evaluation, and approximately 30% of patients are under regular hospital follow-up.12 About 2000 people undergo colectomy for IBD each Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines Table 1 Evidence levels (EL) and recommendation grades (RG) (adapted from the Oxford Centre for Evidence Based Medicine, http://www.cebm.net/index.aspx?o¼1025 (accessed Oct 2010)) EL

Individual study

Technique

1a 1b 1c

Systematic review (SR) with homogeneity of level 1 diagnostic studies Validating cohort studies with good reference standards Specificity is so high that a positive result rules in the diagnosis (SpPIn) or sensitivity is so high that a negative result rules out the diagnosis (SnNout) SR with homogeneity of >level 2 diagnostic studies Exploratory cohort study with good reference standards

Systematic review (SR) with homogeneity of randomised controlled trials (RCTs) Individual RCT (with narrow CI) All or none

2a 2b 2c 3a 3b 4 5 RG A B C D

SR with homogeneity of 3b and better studies Non-consecutive study; or without consistently applied reference standards Caseecontrol study, poor or non-independent reference standard Expert opinion without explicit critical appraisal, or based on physiology, ’bench research’ or ’first principles’ GRADES OF EVIDENCE Consistent level 1 studies Consistent level 2 or 3 studies or extrapolation from level 1 studies Level 4 studies or extrapolation from level 2 or 3 studies Level 5 evidence or troublingly inconsistent or inconclusive studies of any level

year. The lifetime risk for surgery may be as high as 70e80% for Crohn’s disease and 20e30% for ulcerative colitis, depending on disease severity and location.17e21 Costs of caring for patients with IBD in UK hospitals have recently been assessed.22 Lifetime costs for IBD are comparable to a number of major diseases, including heart disease and cancer. This implies a substantial burden of disease and disability that is mirrored by the large amount of academic and commercial activity currently being expended to develop better treatments. Patients ﬁnd symptoms of ulcerative colitis or Crohn’s disease embarrassing and humiliating. IBD can result in loss of education and difﬁculty in gaining employment or insurance. It can cause psychological problems and growth failure or retarded sexual development in young people. Medical treatments (steroids, immunosuppressants) can cause secondary health problems, and surgery may result in complications such as impotence or intestinal failure. The impact of IBD on society is disproportionately high as presentation often occurs at a young age and has the potential to cause lifelong ill health.

2.2 UK IBD service standards Historically, practice guidelines have focused on evidence-based therapeutics. This addresses only one aspect of care. Since the 2004 BSG guidelines were published, two signiﬁcant steps have been taken to address the quality of care provided to patients with IBD in the UK. First, The UK IBD National Audit programme was developed in partnership by the British Society of Gastroenterology, the Association of Coloproctology of Great Britain and Ireland, the National Association for Colitis and Crohn’s Disease (NACC) and the Clinical Effectiveness Unit at the Royal College of Physicians, (http://www.rcplondon.ac.uk/resources/ inﬂammatory-bowel-disease-audit (accessed Feb 2011)). The inaugural audit was performed in 2006. It assessed the structure; organisation, processes and outcomes of care for patients with IBD admitted to UK hospitals and found a wide variation in all aspects across the country. The second round audit in 2008 has demonstrated an improvement in many aspects of basic care such as the provision of specialist wards, availability of IBD nurses, the prescription of prophylactic heparin and the collection of stool specimens for culture Clostridium difﬁcile toxin. However, there remains wide variation, with some key deﬁcits in fundamental aspects of IBD care. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

SR (with homogeneity) of cohort studies Individual cohort study (including low quality RCT; eg, <80% follow-up) ’Outcomes’ research; ecological studies SR with homogeneity of case-control studies Individual caseecontrol study Case series (and poor quality cohort and caseecontrol studies) Expert opinion without explicit critical appraisal, or based on physiology, ’bench research’ or ’first principles’

Second, a Working Group of IBD health professionals (chaired by Richard Driscoll, CEO National Association of Colitis and Crohn’s) assembled in 2007 to develop a Statement of Standards for IBD Healthcare that could be applied across the UK. After widespread consultation the document was published in February 2009 and launched in the UK parliaments and legislative assemblies. It sets out the standards that IBD services should attain, but does not prescribe particular models of service organisations. Full copies of the IBD Standards document can be downloaded at http://www.ibdstandards.org.uk (accessed Oct 2010). The Standards of Care are grouped into six areas, addressing all aspects from high quality clinical care provided by a multidisciplinary IBD team, through to shared care, patient support and empowerment, education of patients and staff, IT support, along with a commitment to research and service development (see table 2). Importantly, the Standards will be assessed in future rounds of the National Audit programme. Furthermore, in England the Healthcare Commission has already adopted several key elements in to the Annual Health Check, to cross-check Hospital Trusts’ declarations against Core Standards. IBD Service improvement tools can be found at http://www.ibdstandards.org.uk.

2.3 Sources of patient education and support Written information about IBD should be provided in outpatient clinics, ward, and endoscopy areas (IBD Standard D1). Patients being considered for surgery should be offered information about their operation, and where possible, the option of talking to patients who have had pouch surgery or a permanent ileostomy. Patients should be offered advice on where additional information may be obtained and help in interpreting information where the need arises. Sources are too many to provide a comprehensive list. The following provide access to both general and more detailed information: < Crohn’s and Colitis UK: http://www.nacc.org.uk < The Crohn’s and Colitis Foundation of America. http:// www.ccfa.org < CORE (Digestive Disorders Foundation): http://www.corecharity.org.uk < British Society of Gastroenterology: http://www.bsg.org.uk < NHS Choices: https://www.nhs.uk (accessed Oct 2010). < NHS Evidence: http://www.evidence.nhs.uk < UK Clinical Research Network portfolio (gastrointestinal): http://public.ukcrn.org.uk (accessed Oct 2010). 573

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines Table 2 Standard

IBD service standards Implementation standard

A: High-quality clinical care A1 The IBD team* A2 Essential supporting services* A3 Multi-disciplinary working* A4 Referral of suspected patients with IBD A5 Access to nutritional support and therapy* A6 Arrangements for use of immunosuppressive and biological therapies A7 Surgery for IBD* A8 Inpatient facilities* A9 Access to diagnostic services A10 Inpatient care* A11 Outpatient care* A12 Arrangements for the care of children and young people who have IBD B: Local delivery of care B1 Arrangements for shared care* C: Maintaining a patient-centred service C1 Information on the IBD service* C2 Rapid access to specialist advice* C3 Supporting patients to exercise choice between treatments C4 Supporting patients to exercise choice between care strategies for outpatient management C5 Involvement of patients in service improvement* D: Patient education and support D1 Provision of information* D2 Education for patients D3 Information about patient organisations D4 Support for patient organisations E: Information technology and audit E1 Register of patients under the care of the IBD service E2 Developing an IBD database E3 Participation in audit F: Evidence-based practice and research F1 Training and education F2 Research F3 Service development *Adopted by Healthcare Commission. IBD, inflammatory bowel disease.

< CICRA (Crohn’s in Childhood Research Association): http://

www.cicra.org/

Recommendation for service delivery

< Multi-disciplinary care (IBD Standard A3)

– The IBD team should have regular meetings to discuss patients with complex needs – Patients should have access to a joint or parallel gastroenterologyesurgical clinic that is held at least monthly in a unit that meets the standards set out in this document. < Patient management (IBD Standard A4) – Local protocols should be developed to facilitate referral of symptomatic patients in whom IBD is suspected. – All patients with IBD who are admitted to hospital should be notiﬁed to the IBD specialist nurse (IBD Standard A10). – Newly diagnosed patients for whom surgery is not an immediate consideration should be transferred to the care of the medical gastroenterology team. – IBD inpatients should, wherever possible, be cared for in a specialist ward area with 24 h access to intensive care facilities on site. – IBD surgery should be undertaken by recognised colorectal surgeons, who are core members of the IBD team, or their supervised trainees, in a unit performing such operations regularly (IBD Standard A7). – All IBD outpatients should have an annual review and basic information recorded. This may be in a hospital/community clinic, or by telephone follow-up, and should be done by a healthcare professional with recognised competence in IBD (Standard A11). – Patients with IBD should have access to a dedicated telephone service supported by an answer-phone, which can provide a response by the end of the next working day (Standard A11). – Patients experiencing a possible relapse of their IBD should have access to specialist review within a maximum of ﬁve working days (Standard A11). – There must be a deﬁned policy and protocol for transitional care of adolescents with IBD (Standard A12).

3.0 INFLAMMATORY BOWEL DISEASE 3.1 Definitions The deﬁnitions and diagnosis of ulcerative colitis and Crohn’s disease are thoroughly reviewed in the ECCO consensus documents.2 6 In particular, the deﬁnitions of ulcerative colitis and Crohn’s disease acknowledge the revised Montreal classiﬁcation

< Hospitals involved in the care of patients with IBD should

model their service as far as possible to meet the IBD Service Standards (EL5, RG D).

IBD Service Standards (service delivery): It is suggested that an IBD service is delivered within the following basic framework: < The IBD team – Patients with IBD should be cared for by a deﬁned IBD team with named personnel comprising gastroenterologists, colorectal surgeons, clinical nurse specialists, a dietician, pharmacist, pathologist and GI radiologist (IBD Standard A1). The roles and responsibilities of an IBD nurse specialist are outlined within Royal College of Nursing guidance, (http://www.rcn.org.uk/__data/assets/pdf_ﬁle/0007/107746/ 003194.pdf (last accessed Oct 2010)). The IBD team should have access to the following essential supporting services: a psychologist/counsellor, rheumatologist, ophthalmologist, dermatologist, obstetrician, nutrition support team, a paediatric gastroenterology clinical network, general practise (IBD Standard A2). 574

Table 3 Definition of ulcerative colitis phenotype according to the Montreal classification23 Maximal extent of inflammation observed at colonoscopy Proctitis Left-sided e extending up to splenic flexure More extensive disease

E1 E2 E3

Table 4 Definition of Crohn’s disease phenotype according to the Montreal classification23 Age of onset

Location

Behaviour

#16 years (A1)

Ileal (L1)

1740 years (A2) >40 years (A3)

Colonic (L2) Ileo-colonic (L3) *Isolated upper GI disease (L4)

Non-stricturing, Non-penetrating (B1) Stricturing (B2) Penetrating (B3) + ‘p’ if peri-anal disease

*L4 is a modifier that can be added to L1 e 3 when concomitant upper gastrointestinal (GI) disease is present.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines which attempts to more accurately characterise the clinical patterns of IBD.23 24 Ulcerative colitis is characterised by diffuse mucosal inﬂammation limited to the colon. It is classiﬁed according to the maximal extent of inﬂammation observed at colonoscopy because this is most clearly related to the risk of complications, including dilatation and cancer. The implications of limited macroscopic disease with extensive microscopic inﬂammation remain unclear. Crohn’s disease is characterised by patchy, transmural inﬂammation, which may affect any part of the gastrointestinal tract. It may be deﬁned by: age of onset, location, or behaviour. About 5% of patients with IBD affecting the colon are unclassiﬁable after considering clinical, radiological, endoscopic and pathological criteria, because they have some features of both conditions. This is now termed as ‘IBD, type unclassiﬁed (IBDU)’. The term ‘indeterminate colitis (IC)’ should be reserved for cases where colectomy has been performed and the pathologist remains unable to classify the disease after a full examination.23 If all patients are characterised in this standard fashion, this should facilitate data collection for an IBD registry and clinical research.

3.2 Clinical features and course of disease17e19 Ulcerative colitis

21 25e27

The cardinal symptom of ulcerative colitis is bloody diarrhoea. Associated symptoms of colicky abdominal pain, urgency, or tenesmus may be present. It is a severe disease that used to carry a high mortality and major morbidity. With modern medical and surgical management, the disease now has a slight excess of mortality in the ﬁrst 2 years after diagnosis, but little subsequent difference from the normal population. However, severe colitis is still a potentially life-threatening illness. The clinical course is marked by exacerbation and remission. About 50% of patients have a relapse in any year. An appreciable minority has frequently relapsing or chronic, continuous disease and overall, 20e30% of patients with pancolitis come to colectomy. After the ﬁrst year approximately 90% of patients are fully capable of work (deﬁned by <1 month off work/year), although signiﬁcant employment problems remain an issue for a minority.

Crohn’s disease Symptoms of Crohn’s disease are more heterogeneous, but typically include abdominal pain, diarrhoea and weight loss. Systemic symptoms of malaise, anorexia, or fever are more common. Crohn’s disease may cause intestinal obstruction due to strictures, ﬁstulae (often perianal) or abscesses. Surgery is not curative and management is directed to minimising the impact of disease. At least 50% of patients may require surgical treatment in the ﬁrst 10 years of disease and approximately 70e80% may require surgery within their lifetime, dependent on the site of the disease. The overall mortality is slightly higher than the normal population and is greatest in the 2 years after diagnosis or in those with upper gastrointestinal disease. The clinical course is also characterised by exacerbations and remission. Crohn’s disease tends to cause greater disability than ulcerative colitis with only 75% of patients fully capable of work in the year after diagnosis and 15% of patients unable to work after 5e10 years of disease. Both ulcerative colitis and Crohn’s colitis are associated with an equivalent increased risk of colonic carcinoma.28e31 Smoking increases the risk of Crohn’s disease, but decreases the risk of ulcerative colitis through unknown mechanisms.32 Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

3.3 Diagnosis and investigation2

6 33

The diagnosis of IBD is conﬁrmed by clinical evaluation and a combination of haematological, endoscopic, histological, or imaging-based investigations. In the case of ulcerative colitis the diagnosis should be made on the basis of clinical suspicion supported by appropriate macroscopic ﬁndings on sigmoidoscopy or colonoscopy, typical histological ﬁndings on biopsy and negative stool examinations for infectious agents. For Crohn’s disease the diagnosis depends on demonstrating focal, asymmetric and often granulomatous inﬂammation but the investigations selected vary according to the presenting manifestations, physical ﬁndings and complications. For all patients, there should be local referral patterns agreed so that patients suspected of having IBD can be referred for rapid consultation and assessment.

3.3.1 History and examination A full history should include recent travel, medication (including antibiotics and non-steroidal anti-inﬂammatory drugs), sexual and vaccination history where relevant. Particular attention should be paid to established risk factors including smoking, family history, previous appendicectomy and recent episodes of infectious gastroenteritis. Details should include the stool frequency and consistency, urgency, rectal bleeding, abdominal pain, malaise, fever, weight loss and symptoms of extra-intestinal (joint, cutaneous and eye) manifestations of IBD. Examination should include general well-being, measurement of weight, calculation of body mass index, pulse rate, blood pressure, temperature, check for anaemia, ﬂuid depletion, abdominal tenderness or distension, palpable masses and perineal examination.

3.3.2 Initial investigations Laboratory investigations should include full blood count, urea and electrolytes, liver function tests and erythrocyte sedimentation rate or C reactive protein, ferritin, transferrin saturation, vitamin B12 and folate. Serological markers such as pANCA, ASCA are present in a signiﬁcant proportion of patients with IBD but there is no evidence base to recommend their use in the diagnosis of IBD. Faecal calprotectin is accurate in detecting colonic inﬂammation and can help identify functional diarrhoea. Microbiological testing for Clostridium difﬁcile toxin, in addition to standard organisms, is increasingly important. C difﬁcile infection has a higher prevalence in patients with IBD through unknown mechanisms, may not be conﬁned to the colon, and is associated with increased mortality. A minimum of four stool samples is required to detect 90% of cases.34 35 Cytomegalovirus (CMV) should be considered in severe or refractory colitis, as reactivation is common in patients with IBD on immunosuppression. Additional tests may be needed for patients who have travelled abroad. Abdominal radiography is essential in the initial assessment of patients with suspected severe IBD: it excludes colonic dilatation and may help assess disease extent in ulcerative colitis or identify proximal constipation. In Crohn’s disease abdominal radiography may give an impression of a mass in the right iliac fossa, or show evidence of small bowel dilatation.

3.3.3 Endoscopy Colonoscopy with multiple biopsies (at least two biopsies from ﬁve sites including the distal ileum and rectum) is the ﬁrst line procedure for diagnosing colitis. It allows classiﬁcation of disease based on endoscopic extent, severity of mucosal disease and histological features. It also allows assessment of suspected stenoses in the distal ileum or colon. 575

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines In acute severe colitis, full colonoscopy is rarely needed36 and may be contraindicated. Phosphate enema prior to sigmoidoscopy is considered safe in acute severe colitis, except in those with colonic dilatation. A rectal biopsy is best taken for histology even if there are no macroscopic changes. Upper gastrointestinal (GI) endoscopy should be considered in coexisting dyspepsia. The role of small bowel endoscopy (enteroscopy/capsule) has yet to be deﬁned. The available evidence is summarised in a consensus statement produced by OMED/ ECCO.37

3.4 Histopathology Histopathological examination of biopsy specimens should be carried out according to the BSG guideline, ‘A Structured Approach to Colorectal Biopsy Assessment’.38 There should be an attempt to deﬁne the type of IBD, to mention other coexistent diagnoses or complications and to mention the absence or presence of any dysplasia and its grade. The appropriate term for IBD that cannot be classiﬁed is ‘IBD Unclassiﬁed’.23 Medical and surgical therapy may modify the histological appearances of IBD and these should be taken into account when assessing IBD biopsy pathology.2 39

3.5 Imaging modalities Imaging can be helpful in diagnosis, assessment of disease extent and severity and for investigation of suspected complications. Each modality has its own advantages and drawbacks and the tests are often complimentary. It is desirable for clinicians to discuss imaging with a radiologist to avoid unnecessary exposure to ionising radiation (see table 5).40 41

3.5.1 Ultrasound Ultrasound cannot comprehensively assess the gut when used in isolation. It is the ﬁrst-line test for gallstones and kidney stones, which should not be forgotten as complications of Crohn’s disease. In expert hands it has a high sensitivity for detecting disease, particularly in the terminal ileum. However, such expertise is not widely available in the UK. Doppler techniques are useful in the assessing the degree of disease activity. It has reasonable sensitivity for documenting the presence of complicating abscess, particularly in thinner patients and is a useful ﬁrst line test in this context.42 43 Importantly, there is no radiation dosage or contrast agent needed and it is safe in pregnancy.

3.5.2 Magnetic resonance imaging Modern MRI hardware and software facilitate rapid and accurate assessment of the small bowel. Importantly, there is no radiation dosage which makes the technique ideally suited to the Crohn’s disease population given their age demographic and need for repeat imaging. Employed sequences are complimentary in characterising the bowel wall in Crohn’s disease (eg, documenting the presence of mural oedema or abnormal gadolinium Table 5 Dosage and risk associated with diagnostic x-ray procedures, www.hpa.org.uk (accessed Oct 2010) Diagnostic x-ray procedure

Typical effective doses (mSv)

Lifetime additional risk of fatal cancer per examination

Chest Pelvis Abdomen Barium follow-through Barium enema CT abdomen/pelvis

0.02 0.7 0.7 3 7 10

1 1 1 1 1 1

576

in in in in in in

1 000 000 30 000 30 000 6700 3000 2000

enhancement patterns). Large comparative trials with conventional ﬂuoroscopic barium techniques are currently lacking but recent data (mainly single-site studies) suggest MRI is equivalent or superior, particularly in those with established disease. Early mucosal disease such as aphthous ulceration is better seen by wireless capsule endoscopy or high-quality ﬂuoroscopic studies.44 45 MRI provides information about disease activity and may be useful in distinguishing between inﬂammatory and ﬁbrotic stricturing. It also has very high sensitivity for detection of extraluminal complications (including abscess formation) and demonstrates internal ﬁstulisation with good accuracy. Pelvic MRI has a particular place in the evaluation of perianal disease, and provides a complementary mode of assessment to endo-anal ultrasound and examination under anaesthetic. MR enterography is more widely performed in the UK than MR enteroclysis. Availability of small bowel MRI (both equipment and expertise) is currently limited to around 40% of UK institutions. Magnetic resonance cholangiopancreatography is the initial investigation of choice in suspected sclerosing cholangitis.

3.5.3 Computed tomography scanning CT imaging of the bowel (either CT enteroclysis or CT enterography) provides similar information to MRI, although tissue characterisation capability is less. It is traditionally the ‘gold standard’ for the detection of extraluminal complications, notably abscess formation. Intravenous contrast administration is usually performed during CT. Advantages over MRI include widespread availability, rapid image acquisition (few seconds) and superior spatial resolution. However CT imparts a signiﬁcant radiation burden, which may carry a cancer risk.41 Furthermore radiation cumulative doses may be signiﬁcant with repeat imaging.46 Provision of CT enterography/enteroclysis is currently similar to MRI in the UK. Unprepared CT (ie, without bowel distension) has an important role in rapidly and accurately assessing patients for acute complications such as obstruction or sepsis. Importantly CT is usually available out of hours.

3.5.4 Barium fluoroscopy High-quality barium studies have superior sensitivity over crosssectional techniques for subtle early mucosal disease, although in those with established and/or more advanced disease, both CT and MR may be equivalent and also provide information on submucosal disease. Barium ﬂuoroscopy imparts a radiation dose to patients (approximately one third to one half of the CT dose) with its associated risks. This risk may be of particular importance to young patients with IBD requiring immunosuppressive therapies.

3.5.5 Isotope-labelled scans A variety of nuclear medical techniques can be used in the assessment of IBD, although they have no role in the primary diagnosis of IBD.47 Technetium-99m labelling of white blood cells remain a widely acceptable scintigraphic method for the evaluation of disease extension and severity. Positron emission tomography alone or with CT using ﬂuorine-18 ﬂuorodeoxyglucose appears to be a promising method of measuring inﬂammation in patients with IBD. These techniques might be considered when colonoscopy is not completed successfully or other imaging modalities are negative.

3.5.6 Imaging common clinical scenarios Suspected severe IBD The plain abdominal x-ray is essential in the initial assessment of patients with suspected severe IBD: it excludes colonic Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines dilatation and may help assess disease extent in ulcerative colitis or identify proximal constipation. In Crohn’s disease abdominal radiography may give an impression of a mass in the right iliac fossa, or show evidence of small bowel dilatation.

Assessing the small bowel in Crohn’s disease

MRI, CT, ultrasound and barium ﬂuoroscopy all have established roles in deﬁning disease extent in those with known disease. The role of small bowel endoscopy (enteroscopy/ capsule) has yet to be deﬁned. The available evidence is summarised in a consensus statement produced by OMED/ ECCO.37 Choice depends on local availability and expertise as well as patient factors and clinical indication. However, whenever possible, those techniques not imparting radiation should be employed. Small studies in expert centres have suggested MRI and CT have the better diagnostic yield partly because of the detection of extra mucosal disease.45 48 49

Imaging and anti-TNF therapy in Crohn’s disease Before using anti-TNF therapy it is important to exclude un-drained abscess collections where these may be present. This is particularly true in ﬁstulising disease. Cross-sectional techniques including ultrasound, MRI or CT should be employed. CT and MRI are overall more sensitive than ultrasound. MRI may be used to monitor response of ﬁstulising perianal disease.

Recommendations

< All patients with diarrhoea should have stools sampled for

culture and C difﬁcile toxin. Four samples are required for 90% sensitivity. (EL4, RGC). < Imaging techniques may be constrained by availability and local expertise. In general, attempts should be made to minimise exposure to ionising radiation. < For imaging the small bowel, MRI is the preferred technique where available (EL 2b). < All new patients should have their disease phenotype classiﬁed in accordance with the Montreal Classiﬁcation (EL 5, RG D).

IBD Service Standards for diagnosis and investigation: (IBD standard A)

< Local guidelines/referral pathways should be in place for rapid

referral of new/suspected cases of IBD. < The patient’s weight and body mass index (BMI) should be

measured at each attendance. < Outpatients should wait no more than 4 weeks for radiological/ endoscopic investigations. < Inpatients with severe disease should wait no more than 24 h for necessary imaging or endoscopy. < Processing of biopsies should be rapid (2e5 days maximum according to need).

4.0 THERAPEUTIC OPTIONS IN THE MANAGEMENT OF IBD http://www.bsg.org.uk/forum (accessed Oct 2010)/(This section examines data on efﬁcacy: speciﬁc recommendations are included in sections 5, 6, and 7).

4.1 Nutrition Malnutrition in IBD is common and multi-factorial in origin. Nutritional assessment, including BMI is important: there are validated tools such as Malnutrition Universal Screening Tool (MUST) to guide assessment,50 (http://www.bapen.org.uk/ musttoolkit.html (last accessed Oct 2010)). Patients with active colitis may have secondary lactose intolerance and a dairy free diet may reduce gas and bloating (EL5, RGD). Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

4.1.1 Nutritional support, (http://www.nice.org.uk/nicemedia/pdf/cg032fullguideline.pdf (last accessed Oct 2010)) Micronutrients Speciﬁc attention should be paid to nutrient deﬁcits such as calcium, vitamin D, other fat soluble vitamins, zinc, iron and (after ileal resection especially) vitamin B12 status. Serum vitamin B12 is best measured annually in patients with ileal Crohn’s disease.51

Macronutrients In speciﬁc circumstances, protein and caloric support is indicated, such as when the absorptive capacity of the gut is reduced in short bowel syndrome or in the perioperative care of patients with signiﬁcant (more than 15%) weight loss or low BMI.52 This may mean total parenteral nutrition (TPN) including home TPN in a minority of Crohn’s disease patients with intestinal failure. Approximately 20% of the home TPN patients in Europe have underlying Crohn’s disease that is about one case per 1.5 million population.53 See BSG guidelines on short bowel syndrome,54 (http://www.bsg.org.uk/clinical-guidelines/smallbowel-nutrition/guidelines-for-management-of-patients-with-ashort-bowel.html (last accessed Oct 2010)).

4.1.2 Nutritional therapy55 Therapeutic liquid feeds There is no indication for liquid feed to treat ulcerative colitis. Enteral nutritional therapy alters the inﬂammatory response in Crohn’s disease and may be useful in therapy.56 57 In Crohn’s disease, exclusive enteral nutrition (EEN), usually given for 3e6 weeks, is an alternative therapy to corticosteroids for active Crohn’s disease. There is no difference in efﬁcacy between elemental and polymeric diets when used to induce remission in Crohn’s disease.58 When used in children EEN is effective at inducing remission for small and large bowel disease in 60e80%. However, in adults liquid feeds appear less effective than corticosteroids in controlled studies (EL2b) although this may relate to tolerability. The efﬁcacy of EEN to treat active Crohn’s disease has not been assessed in controlled studies against normal diet. There is little evidence to support the use of liquid feeds as maintenance therapy for Crohn’s disease.59

Advantages of exclusive enteral nutrition Liquid feeding as an alternative to steroids may avoid adverse effects of steroids and ensure optimal growth before fusion of epiphyses prohibits further growth.58 60 Against this is the issue of tolerability in adults of feeds taken orally.

Prebiotics Prebiotics are non-digestible dietary carbohydrates,such as fructo-oligosaccharides which are fermented by the gut microﬂora to produce short-chain fatty acids. Their role is unproven to date.61

Probiotics Bacteria or yeast generally ingested orally as therapy are termed probiotics. They may be administered as a single organism or a deﬁned mixture, aiming to beneﬁcially alter the microbial ecology of the gut. The agents most studied in IBD are E coli Nissle 1917, VSL#3, Lactobacillus rhamnosius GC, Biﬁdobacterium and Saccharomyces boulardii.62 There is evidence for the effectiveness of VSL#3 in preventing pouchitis63 64 (see section 5.7) and some evidence of beneﬁt in maintenance and treatment of ulcerative colitis.65 66 Three randomised placebo-controlled studies have shown that E coli Nissle 1917 (MutaﬂorÒ) 200 mg daily is equivalent to 577

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines standard doses of mesalazine in maintaining remission in ulcerative colitis67 (EL 1b, RG A), and therefore may be an option for patients who are unable or unwilling to take mesalazine. There is no clear evidence to support any role of probiotics in the maintenance of Crohn’s disease either after surgical or medically induced remission.68

Total parenteral nutrition with ‘bowel rest’69e72 There is no good evidence to support the use of parenteral nutrition as an adjunct or sole therapy to induce remission in Crohn’s disease or ulcerative colitis.

4.2 Smoking cessation32

73e78

Smoking is an important environmental factor in the pathogenesis of IBD, though the mechanisms remain under investigation. Current smokers are more likely to develop Crohn’s disease and, following diagnosis, have a poorer prognosis with a signiﬁcantly higher chance of surgical resection, and (if smoking still continues) a greater chance of recurrence at the surgical anastomosis. Smoking cessation is associated with a 65% reduction in the risk of a relapse as compared with continued smokers, a similar magnitude to that obtained with immunosuppressive therapy.79

4.3 Non-steroidal anti-inflammatory drugs There are many publications claiming an adverse effect of non-steroidal anti-inﬂammatory drugs (NSAIDs) in precipiating de novo IBD or exacerbating pre-existing disease, although the evidence remains contradictory and confusing. The current position is summarised in a recent review.80 There is some evidence that mucosal damage is mediated by dual inhibition of COX-1 and COX-2. Selective inhibition with COX-2 inhibitors or COX-1 inhibition with low dose aspirin seems to be safe, at least in the short term.81

4.4 Drugs used in the treatment of inflammatory bowel disease Therapy for IBD is a rapidly evolving ﬁeld, with many new biological agents under investigation that are likely to change therapeutic strategies radically in the next decade. Details of the principal drugs can only be summarised in this document. Patient information sheets can be downloaded from: http:// www.bsg.org.uk/patients/general/patient-information.html (last accessed Oct 2010).

4.4.1 Aminosalicylates82 5-Aminosalicylic acid (5-ASA) or mesalazine (‘mesalamine’ in the USA) can be delivered in millimolar concentrations to the gut lumen by a variety of oral tablets, sachets or suspensions using pH-dependent release mechanisms, multimatrix delivery systems, or conjugation via a diazo bond to a variety of carrier molecules with release of 5-ASA after splitting by bacterial enzymes in the large intestine. They can also be used as topical agents in the form of liquid or foam enemas, or suppositories. They act on epithelial cells by a variety of mechanisms to moderate the release of lipid mediators, inﬂammatory cells, cytokines and reactive oxygen species. Oral forms include: e pH-dependent release/resin coated (AsacolÒ, SalofalkÒ, IpocolÒ, MesrenÒ) e time-controlled release (PentasaÒ) e Multimatrix delivery systems (Mezavant XLÒ) e delivery by carrier molecules, with release of 5-ASA after splitting by bacterial enzymes in the large intestine (sulfasalazine (SalazopyrinÒ), olsalazine (DipentumÒ), balsalazide (ColazideÒ)). 578

Efficacy in ulcerative colitis83

84

For ulcerative colitis, greater clinical improvement (but not necessarily remission) is associated with doses >3 g/day. Clinical improvement characteristically occurs at twice the remission rate. In a meta-analysis of oral 5-ASA for active ulcerative colitis, of 19 trials involving 2032 patients, nine were placebo controlled and 10 compared mesalazine with sulfasalazine. The outcome of interest on an intention-to-treat principle was the failure to induce remission, so that a pooled OR <1.0 indicates one treatment to be more effective than another. Mesalazine was more than twice as effective as placebo (OR 0.39; CI 0.29 to 0.52, but not signiﬁcantly better than sulfasalazine (OR 0.87; CI 0.63 to 1.20). More recent trials have studied the efﬁcacy of high dose 5-ASA in ulcerative colitis. The rate of remission at the end of these studies is similar on 2.4 g and 4.8 g daily. However, there appeared to be faster resolution of symptoms on 4.8 g daily compared to 2.4 g daily.85 86 Increasing colonic concentrations of 5-ASA by using a combination of oral and topical preparations of mesalazine was shown to be more effective than oral therapy alone even in patients with extensive disease.87 Trials in the acute and maintenance phase (see below) suggest that once daily dosing is effective with most preparations of 5-ASA.85 88 A recent meta-analysis demonstrates that rectal 5-ASA is superior to rectal steroids for the induction of remission of mild-moderate distal ulcerative colitis.89 The main role for 5-ASA is maintenance of remission in ulcerative colitis. All 5-ASA derivatives show comparable efﬁcacy to sulfasalazine, but in a meta-analysis sulfasalazine had a modest therapeutic advantage for maintaining remission (OR 1.29, CI 1.08 to 1.57).90 The choice of 5-ASA is debated, but is inﬂuenced by tolerability (mesalazine is tolerated by 80% of those unable to tolerate sulfasalazine), dose schedule (single- or twice-daily dosing is associated with better compliance) and cost. There is now robust evidence to suggest that single daily dosing is as effective as multiple dosing, and may even be superior.91 92 Efﬁcacy may depend more on adherence with the prescribed dose than the delivery system. If the delivery system is considered important, then the drug is best matched to the site of disease, by using azo-bonded compounds for distal disease. Maintenance therapy with all 5-ASA drugs may reduce the risk of colorectal cancer by up to 75% (OR 0.25, CI 0.13 to 0.48).93 This favours long-term treatment for patients with extensive ulcerative colitis.

Efficacy in Crohn’s disease84 In active Crohn’s ileocolitis, a meta-analysis of the three placebo-controlled trials of Pentasa 4 g daily for 16 weeks in a total of 615 patients, showed a mean reduction of the Crohn’s disease activity index (CDAI) from baseline of e63 points, compared to e45 points for placebo (p¼0.04).94 While this conﬁrms that Pentasa 4 g/day is superior to placebo in reducing CDAI, this is unlikely to be of clinical signiﬁcance. Subgroup analyses do not provide sufﬁciently clear answers to whether one group of patients beneﬁt more than another, and the use of aminosalicylates as ﬁrst line therapy in this group is not justiﬁed by the evidence. The national Cooperative Crohn’s Disease study did identify some beneﬁt in colonic Crohn’s disease from sulfasalazine at a dose of 4e6 g daily, although this was modest. This effect was not seen with newer preparations of 5-ASA in more recent studies.95e101 There is no evidence that 5-ASA is superior to placebo for the maintenance of medically induced remission.102 There is evidence to suggest that mesalazine >2 g/day has a modest effect in reducing relapse after surgery (NNT 10e12), especially after Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines small bowel resection (40% reduction at 18 months). However the cost of therapy and the pill burden for patients are such that it should probably be reserved for speciﬁc patientsdnormally those with small-bowel disease, and only after more effective measures such as smoking cessation have been instituted.84

Adverse effects of 5-ASA103e105 Side effects of sulfasalazine occur in 10e45% of patients, depending on the dose. Headache, nausea, epigastric pain, diarrhoea, and oligospermia in men are most common and dose related. Serious idiosyncratic reactions (including Stevense Johnson syndrome, pancreatitis, agranulocytosis, or alveolitis) are rare. Mesalazine intolerance occurs in up to 15% of patients. Diarrhoea (3%), headache (2%), nausea (2%) and rash (1%) are reported, but a systematic review has conﬁrmed that all new 5-ASA agents are safe, with adverse events that are similar to placebo for mesalazine or olsalazine. No comparison between balsalazide and placebo has been published, but events were lower than with sulfasalazine. Acute intolerance to all 5-ASAs in 3% may resemble a ﬂare of colitis since it includes bloody diarrhoea. Recurrence on rechallenge provides the clue. All aminosalicylates have been associated with nephrotoxicity (including interstitial nephritis and nephrotic syndrome), which appears both to be idiosyncratic and, in part, dose related.106 Reactions are rare, but patients with pre-existing renal disease are at higher risk. A population-based study found the risk (OR 1.60, CI 1.14 to 2.26 compared to normal) to be associated with disease severity rather than the dose or type of mesalazine. For patients on maintenance 5-ASA, annual measurement of creatinine is sensible, although there is no evidence that monitoring is necessary or effective at preventing renal impairment. Aminosalicylates should be stopped if renal function deteriorates.

4.4.2 Antibiotics Antibiotics have an important role in treating secondary complications in IBD, such as abscess and bacterial overgrowth.107 There is some evidence that metronidazole and ciproﬂoxacin have speciﬁc uses in Crohn’s disease. There is no clear-cut evidence to support the use of these antibiotics in ulcerative colitis as disease modifying therapy.

Metronidazole Metronidazole in a synthetic nitroimidazole antibiotic and antiprotozoal drug.

Efficacy in Crohn’s disease. Early trials of metronidazole to treat Crohn’s disease showed reductions in blood markers of inﬂammation (ESR and orosomucoid levels).108e110 Following ileocaecal resection, 20 mg/kg/day for 3 months reduces the risk of endoscopic recurrence in the short term only.111 (see section 6.6.4) Metronidazole is used to treat perianal disease. The evidence is from case series, it has not been subject to adequately powered controlled studies and recent data suggests it is less effective than ciproﬂoxacin.112 Metronidazole treatment of pouchitis improves diarrhoea without a clear effect on pouch inﬂammation.113 It appears less effective than ciproﬂoxacin. Adverse effects. Metronidazole should be used with caution in the long term as peripheral neuropathy can occur after a mean duration of 6 months.114 Ciprofloxacin

This drug is a ﬂuoroquinolone antibiotic with a broad spectrum of activity against Gram positive and negative bacteria including many enteric pathogens. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Efficacy in Crohn’s disease. There are no placebo-controlled studies of ciproﬂoxacin to treat active Crohn’s disease though comparisons have been made with other drugs showing similar response rates to mesalazine and steroids respectively.115 116 Studies in ulcerative colitis show weak or no effect.117 118 The evidence suggests a greater beneﬁt than metronidazole in perianal Crohn’s disease and pouchitis.112 119 Adverse effects. Ciproﬂoxacin is recognised to cause tendon weakness and this effect may be aggravated by steroids.

Antituberculous chemotherapy and other antibiotics in Crohn’s disease These drugs are not discussed here. The reader is referred to the ECCO consensus statement.3

4.4.3 Corticosteroids Corticosteroids are used in the form of oral prednisolone, prednisone, budesonide (among others), or intravenous hydrocortisone and methylprednisolone. Topical suppositories, foam or liquid enemas include hydrocortisone, prednisolone metasulfobenzoate, betamethasone and budesonide. Many strategies attempt to maximise topical effects while limiting systemic side-effects of steroids. Budesonide (EntocortÒ, BudenofalkÒ) is a poorly absorbed corticosteroid with limited bioavailability and extensive ﬁrst-pass metabolism that has therapeutic beneﬁt with reduced systemic toxicity in ileo-caecal Crohn’s disease, or ulcerative colitis.120 Beclometasone dipropionate has been studied in oral and enema forms in ulcerative colitis, and is no better than 5-ASA.121 122

Choice and mechanism123 Corticosteroids are potent anti-inﬂammatory agents for moderate to severe relapses of both ulcerative colitis and Crohn’s disease. They have no role in maintenance therapy for either disease. They act through inhibition of several inﬂammatory pathways: suppressing interleukin transcription, induction of Ikb that stabilises the NFkb complex, suppression of arachidonic acid metabolism and stimulation of apoptosis of lymphocytes within the lamina propria of the gut. The anti-inﬂammatory dose equivalence of prednisolone 5 mg is betamethasone 0.75 mg, methylprednisolone 4 mg and hydrocortisone 20 mg though with differing mineralocorticoid effects (British National Formulary, http://bnf.org/bnf/index.htm (accessed Oct 2010)).

Efficacy for active ulcerative colitis124e126 Oral prednisolone (starting at 40 mg daily) induced remission in 77% of 118 patients with mild-to-moderate disease within 2 weeks, compared to 48% treated with 8 g/day sulfasalazine. A combination of oral and rectal steroids is better than either alone. Adverse events are signiﬁcantly more frequent at a dose of 60 mg/day compared to 40 mg/day, without added beneﬁt, so 40 mg appears optimal for outpatient management of acute ulcerative colitis. Too rapid reduction in the dose of steroids can be associated with early relapse and doses of prednisolone <15 mg day are ineffective for active disease. Budesonide (colonic release preparation) appears as effective as prednisolone for mildemoderate left-sided and extensive colitis though in a different formulation to that available for Crohns disease.127 Rectal steroids are effective additional treatments in addition to oral salicylates in mild distal disease, but appear less effective than topical aminosalicylates.128 129

Efficacy for active Crohn’s disease95

130e132

Two major trials established corticosteroids as effective therapy for inducing remission in Crohn’s disease. The National 579

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines Co-operative Crohn’s disease Study randomised 162 patients, achieving 60% remission with 0.5e0.75 mg/kg/day prednisone (the higher dose for more severe disease) and tapering over 17 weeks, compared to 30% on placebo (NNT¼3). The comparable European Cooperative Crohn’s Disease Study on 105 patients achieved 83% remission on prednisone 1 mg/kg/day compared to 38% on placebo (NNT¼2) over 18 weeks. The high placebo response rate should be noted, because disease activity in Crohn’s disease (and ulcerative colitis) ﬂuctuates spontaneously. No formal doseeresponse trial has been performed, but 92% remission within 7 weeks was achieved in 142 patients with moderately active Crohn’s disease given prednisone 1 mg/kg/day with no tapering. Unfortunately, the majority of patients do not remain in remission following a ﬁrst dose of steroids; at 1 year, a prolonged steroid response occurs in 44%, with steroid dependency in 36%, and steroid resistance in 20%.133 Budesonide is slightly less effective than prednisolone, but is an appropriate alternative for active ileo-ascending colonic disease. Although steroid therapy provides a symptomatic response in the short term and may induce symptomatic remission, this is not typically associated mucosal healing.134 135

Deciding to treat with steroids132 Efﬁcacy should be balanced against side effects, but decisive treatment of active disease in conjunction with a strategy for complete withdrawal of steroids, is often appreciated by a patient suffering miserable symptoms. Regimens of steroid therapy vary between centres. There is no evidence to support any particular regimen. Two commonly used regimens are: < A starting dose of 40 mg prednisolone per day, reducing by 5 mg/d at weekly intervals, or (for moderate disease). < 20 mg/d for 4 weeks then reduce by 5 mg/day at weekly intervals. A standard weaning strategy helps identify patients who relapse rapidly or do not respond and need adjunctive therapy with thiopurines or as an inpatient. Steroid resistance or unresponsiveness should lead to escalation of treatment, or consideration of surgery. Medical therapies include an immunosuppressive appropriate to the acuteness and type of disease (typically thiopurine in moderate ulcerative colitis or Crohn’s disease, anti-TNF therapy in Crohn’s disease and ciclosporin (or inﬂiximab if ciclosporin is contraindicated) in acute severe ulcerative colitis). Escalation of therapy should be considered in the following situations: < any patient who has a severe relapse or frequently relapsing disease < those who require two or more corticosteroid courses within a 12 month period < those whose disease relapses as the dose of steroid is reduced below 15 mg < relapse within 6 weeks of stopping corticosteroids

Adverse effects of steroids Three broad groups can be identiﬁed, although 50% of patients report no adverse event. < Effects due to supra-physiological doses include cosmetic (acne, moon face, oedema), sleep and mood disturbance, dyspepsia or glucose intolerance. The uncontrolled observational data from the large TREAT registry suggests a twofold RR of infection associated with steroid usage versus no steroid usage and a twofold RR of mortality with prednisolone. Steroids are also associated with increased risk of infections following surgery (OR 1.68 (1.24 to 2.28)).136 137 < Effects associated with prolonged use (usually >12 weeks, but sometimes less) include posterior subcapsular cataracts, 580

osteoporosis, osteonecrosis of the femoral head, myopathy and susceptibility to infection. Steroids have been associated with impaired growth velocity in some conditions. However, when strategies are taken to avoid steroids in Crohn’s disease, the main inﬂuence on growth velocity is disease activity.138 < Effects during withdrawal include acute adrenal insufﬁciency (from sudden cessation), corticosteroid withdrawal syndromea syndrome of myalgia, malaise and arthralgia (similar to recrudesence of Crohn’s disease), or raised intracranial pressure.

Monitoring for side effects Other guidelines recommend monitoring for eye, bone and other side effects particularly in patients on steroids for more than 3 months139 (see also section 7.6: Osteoporosis).

4.4.4 Thiopurines Azathioprine (AZA) or mercaptopurine (MP) are widely used in ulcerative colitis and Crohn’s disease as adjunctive therapy and as corticosteroid-sparing therapies although they are unlicensed therapies for IBD. Their slow onset of action precludes usage as sole therapy for active disease. Purine antimetabolites inhibit ribonucleotide synthesis, but the mechanism of immunomodulation is by inducing T cell apoptosis by modulating cell (Rac1) signalling.140 AZA is non-enzymatically metabolised to MP, which involves loss of a nitro-imidazole side chain; this is thought to explain some of the side effects seen with AZA and which may be less of a problem with MP.141 142 MP is subsequently metabolised to 6-thioguanine nucleotides (6-TGN). 6-TGN has been used for treatment of IBD, but caution is appropriate because of potential hepatotoxicity.

Efficacy in ulcerative colitis AZA is more effective than mesalazine at induction of clinical and endoscopic remission in steroid dependent ulcerative colitis143 and should be ﬁrst-choice therapy in this situation providing other causes of persistent symptoms such as cytomegalovirus or cancer have been excluded. Thiopurines are effective maintenance therapy for patients with ulcerative colitis who have failed or who cannot tolerate mesalazine and for patients who require repeated courses of steroids, although the data quality has been cited as poor in a recent Cochrane review144 and the evidence for using thiopurines in ulcerative colitis is weaker than in Crohn’s disease: probably the best study to date is Ardizzone et al143 which found steroid-free, clinical and endoscopic remission in 53% patients on AZA compared with 21% given 5-ASA (OR on ITT 4.78, 95% CI 1.57 to 14.5).

Efficacy in Crohn’s disease Thiopurines are effective for both induction and maintenance of remission in Crohn’s disease. A Cochrane review of the efﬁcacy of AZA and MP for inducing remission in active Crohn’s disease demonstrated a beneﬁt for thiopurine therapy compared to placebo with an OR of 2.43 (95% CI 1.62 to 3.64). This equates to a number needed to treat (NNT) of about ﬁve and a number needed to harm (NNH) of 14.145 Their efﬁcacy at maintaining remission is conﬁrmed in another Cochrane review. The OR for maintenance of remission with AZA was 2.32 (95% CI 1.55 to 3.49) with a NNT of six. The OR for maintenance of remission with MP was 3.32 (95% CI 1.40 to 7.87) with a NNT of four. Higher doses of AZA improved response. Withdrawals due to adverse events were more common in patients treated with AZA (OR 3.74; 95% CI 1.48 to 9.45, NNH¼20) than with placebo.146 For those who relapse once immunosuppressants are stopped, current evidence suggests that AZA/MP can be safely restarted Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines and continued. The efﬁcacy of thiopurines for post-operative prophylaxis of Crohn’s disease is discussed in section 6.6.4.

Dosing Tailoring or optimisation of thiopurine therapy can occur prior to or during treatment. The appropriate maintenance dose of AZA is 2e2.5 mg/kg/day and of MP is 0.75e1.5 mg/kg/day in both ulcerative colitis and Crohn’s disease. The ‘maximum’ dose will differ between individuals and effectively means that level at which leucopenia develops. There is some evidence that mesalazine has synergistic effects on thiopurine therapy but the mechanism of this effect is unclear.147e149

Is measurement of thiopurine methyl-transferase necessary? AZA induced myelosuppression linked to thiopurine methyltransferase (TPMT) deﬁciency and elevation of thioguanine nucleotide cytotoxic metabolites has been documented in many patient groups including those with IBD.150 TPMT activity in human tissues is under the control of a common genetic polymorphism. About 90% of the population have normal or high enzyme activity and are homozygous for the wild-type allele, 10% inherit intermediate levels of enzyme activity with one wild-type and one variant allele, and one in 300 subjects have no functional TPMT activity. Patients with leukaemia who are TPMT deﬁcient are at increased risk of myelotoxicity. This does not necessarily apply in IBD; in one study the majority (77%) of 41 patients with IBD with AZA-induced bone marrow suppression did not carry a TPMT mutation. Evidence that TPMT activity predicts other side effects or outcome is limited.151 Patients with high levels of TPMT may convert the majority of 6-MP into 6-MMP with inadequate production of 6-TGNs to provide therapeutic efﬁcacy. The precise role of measuring TPMT levels in starting AZA/ MP therapy is still controversial. At the start of AZA/MP therapy, measuring TPMT has a role in identifying the one in 300 patients at risk of severe immunosuppression when treated with standard doses. Most patients who develop leucopenia will have a normal TPMT. During the initial months of AZA therapy a knowledge of low TPMT activity warns of possible early bone marrow toxicity (probability of myelotoxicity in high TPMT group is 3.5% compared 14.3% in the TPMT intermediate group.152 In patients established on AZA, there is no good evidence to suggest that TPMT is predictive of clinical response or drug toxicity, suggesting a role for TPMT in the prediction of early events rather than long-term control.153

Risk of malignancy Organ transplant recipients who are prescribed thiopurines as part of their immunosuppression are recognised to have an increased risk of developing lymphoproliferative disorders.158 In IBD, large population-based studies have shown no increased risk.159 160 However, the data from studies examining patients prescribed thiopurines has been conﬂicting. One meta-analysis suggested no increased risk of malignancy,161 whereas a second suggested a fourfold increased risk of lymphoma in patients with IBD treated with AZA/MP compared with background population.162 More recently, a large prospective study followed almost 20 000 consecutive patients over a 3-year period for the incidence of lymphoproliferative disorders. Those patients receiving maintenance thiopurines had a ﬁvefold increased risk compared to those who had previously or never received the drug.163 In absolute terms, the risk remains very small (<1% risk after 10 years of thiopurine use) and the beneﬁts of AZA outweigh any risks.164 The risk of lymphoma when a thiopurine is combined with Anti-TNF therapy is discussed later (see section 4.4.7). There is an increased risk of non-melanoma skin cancer in patients treated with thiopurines. Patients should be advised to avoid excessive sun exposure and use a high-strength sun block.165

Risk of postoperative complications in patients on thiopurines

Monitoring thiopurine therapy Manufacturers recommend weekly full blood counts (FBCs) for the ﬁrst 8 weeks of therapy followed by blood tests at least every 3 months. There is no evidence that this is effective. One fairly common practice is to perform a full blood count every 2e4 weeks for 2 months and then every 4e8 weeks. The rationale for this approach is that, of patients who develop thiopurine-associated myelotoxicity, approximately half will develop it within 2 months and nearly two thirds within 4 months.154 The mean corpuscular volume is expected to rise on thiopurines and can be used as a surrogate marker for rising 6-TGN concentrations.155

Adverse effects of thiopurines145

between blood tests, although it is rare (around 3% in a review of 66 studies).156 Bone marrow toxicity has been reported to occur up to 11 years after starting AZA157 and blood monitoring should continue throughout thiopurine therapy. Hepatotoxicity and pancreatitis are uncommon (<5%). Patients should be advised to report promptly if a sore throat or any other evidence of infection occurs. Although a signiﬁcant proportion of patients experience adverse effects with thiopurines (28% of 622 patients experienced side-effects in a recent Cochrane review) when the drug is tolerated for 3 weeks, long-term beneﬁt can be expected. Thiopurines can reasonably be continued during pregnancy if ulcerative colitis or Crohn’s disease has been refractory. In a study of 155 men and women with IBD who were parents of 347 pregnancies while taking MP there was no difference in miscarriage, congenital abnormality or infection rate in the thiopurine group compared to a control group. Thiopurine doses should be reduced in renal impairment. The effect and toxicity of AZA/MP is increased by concurrent administration of allopurinol, and we suggest co-prescription be avoided.

146

Adverse events occur in up to 20%. The commonest are allergic reactions (fever, arthralgia, rash) that characteristically occur after 2e3 weeks and cease rapidly when the drug is withdrawn. Profound leucopenia can develop suddenly and unpredictably, in Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Available evidence generally does not suggest an increased rate of postoperative complications associated with immunosuppressive use166 although there is one study which reported an association with postoperative intra-abdominal septic complications.167

Duration of maintenance therapy with thiopurines Recent evidence favours indeﬁnite use of AZA/MP once remission has been established. Fraser et al carried out a retrospective study of 622 patients (272 Crohn’s disease, 346 ulcerative colitis) at a single centre treated over 30 years with AZA,168 and found a 60e75% relapse rate 3 years after stopping immunosuppressants with no effect of duration of AZA dosage. Lémann and colleagues in GETAID carried out a randomised controlled study of AZA withdrawal in patients with Crohn’s disease in longterm remission on AZA.169 Median durations of AZA therapy and of clinical remission were 68.4 months and 63.6 months 581

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines respectively. They found an 8% relapse rate in those randomised to continue AZA therapy compared with 21% relapse rate in those randomised to stop AZA, with high CRP (>20 mg/l) and Hb<12 being associated with increased RR. Subsequent followup of the cohort who stopped AZA has conﬁrmed a high relapse rate, with 14%, 53% and 63% relapsing at 1, 3 and 5 years respectively.170 Kim et al looked at 120 patients who were treated with MP for at least 6 months, achieved remission within 1 year of therapy, and who were in prolonged clinical remission without steroids (> 6 mo without steroids).171 For 84 patients maintained on MP, cumulative relapse rates at 1, 2, 3, and 5 years were 29%, 45%, 55%, and 61%. For 36 patients who stopped MP, relapse rates at the same intervals were 36%, 71%, 85% and 85%. Sex, disease distribution, disease duration, time to remission on MP and concomitant 5ASA use did not affect relapse rates. Cassinotti et al found for ulcerative colitis that disease extent, lack of sustained remission during AZA, and duration of therapy may stratify the risk of relapse on AZA withdrawal.172

4.4.5 Methotrexate Polyglutamated metabolites of methotrexate (MTX) inhibit dihydrofolate reductase, but this cytotoxic effect does not explain its anti-inﬂammatory effect. Inhibition of cytokine and eicosanoid synthesis probably plays a role. At present MTX is positioned as a second-line immunosuppressive agent in patients resistant or intolerant of AZA or MP, although it is currently unclear whether thiopurines are any more efﬁcacious than MTX for induction or maintenance of remission in IBD.

Efficacy in Crohn’s disease

MTX is effective for the induction173 and maintenance174 of remission in Crohn’s disease and may induce mucosal healing.175 176 Evidence from a single large RCT of adult patients demonstrates that 25 mg/week of intramuscular MTX is more effective than placebo at inducing steroid-free remission at 16 weeks (39% vs 19%; p¼0.025; NNT¼5).177 In a subsequent study patients who responded to induction therapy were randomised to 15 mg/week of intramuscular MTX. 65% of patients in the treated group compared with 39% in the placebo group were in remission at 40 weeks (NNT¼4).178

Efficacy in ulcerative colitis No comparable trials have addressed the role of MTX in the induction or maintenance of remission in ulcerative colitis. A single RCT of low dose (12.5 mg once weekly) oral MTX was not shown to be efﬁcacious at inducing or maintaining remission179 and is the only study considered in the Cochrane review.180 The low dose and oral administration may account for the disappointing response. Several retrospective series, using larger weekly doses, have published more promising data with response or remission rates up to 78% in patients with ulcerative colitis resistant or intolerant of AZA or MP.181e184

Mode of delivery Parenteral administration (either subcutaneous or intramuscular) may be more effective that oral therapy and is recommended, although oral dosing may be more convenient. Studies in rheumatoid arthritis indicate the bioavailability of intramuscular MTX is greater than oral administration and equivalent to subcutaneous dosing.185 Consistent with this, a large RCT in rheumatoid arthritis demonstrated subcutaneous MTX is significantly more effective than oral administration.186 To date there is no comparable studies in IBD; however, small uncontrolled 582

series indicate that parenteral might be superior to oral administration in maintaining remission187 188 and subcutaneous may be as effective as intramuscular dosing.189

Monitoring therapy Measurement of full blood count and liver function tests are advisable before and within 4 weeks of starting therapy, then monthly. The same caveats as for monitoring thiopurine therapy apply. Patients should remain under specialist follow-up.

Adverse effects of methotrexate190

191

Side effects are reported by 27e49% of patients leading to drug discontinuation in 10e25% of MTX treated patients. Early toxicity from MTX is primarily gastrointestinal (nausea). Co-prescription of folic acid 5 mg (once a week, taken 3 days after MTX) limits GI side effects of nausea, vomiting, diarrhoea and stomatitis. Long-term concerns are hepatotoxicity, pneumonitis and opportunistic infections. A study of liver biopsies in patients with IBD taking MTX showed mostly only mild histological abnormalities, despite cumulative doses of up to 5410 mg. Hepatotoxicity may be minimised by avoiding administration in patients with signiﬁcant alcohol consumption, type II diabetes, obesity and concurrent liver diseases which may cause steatohepatitis. Surveillance liver biopsy is not warranted, but if the alanine aminotransferase (ALT) doubles then it is sensible to withhold MTX until it returns to normal before a rechallenge. The prevalence of pneumonitis has been estimated to be two to three cases per 100 patient-years of exposure, but large series have not reported any cases. MTX is teratogenic and should not be used in women or men considering conception. It may persist in tissues for long periods; therefore conception should be avoided for 3e6 months after withdrawal of therapy. Breast-feeding is not recommended.192

Duration of therapy Evidence regarding duration of treatment with MTX is lacking and no recommendation can be given. A meta-analysis of observational studies reports remission rates of 75%, 53% and 43% after 1, 2 and 3 years of treatment respectively.193 Prolonged use may be considered if needed. Potential risks and beneﬁts should be discussed on an individual basis.

4.4.6 Calcineurin inhibitors Ciclosporin (oral or intravenous, unlicensed therapy for ulcerative colitis) Ciclosporin (CsA) is an inhibitor of calcineurin, which prevents clonal expansion of T cell subsets. It has a rapid onset of action and is effective in the management of severe ulcerative colitis.

Efficacy in ulcerative colitis Intravenous CsA is rapidly effective as a salvage therapy for patients with refractory ulcerative colitis, who would otherwise face colectomy, but its use is controversial because of toxicity and long-term failure rate. The drug should rarely be continued for more than 3e6 months and its main role is a bridge to thiopurine therapy (see section 4.4.4). However, a Cochrane review has concluded that numbers in controlled trials are so few (only 50)194 195 that there was limited evidence for CsA being more effective than standard treatment alone for severe ulcerative colitis.196 At the time of writing there are two large ongoing controlled trials comparing CsA with inﬂiximab in the treatment of acute severe colitis.

Efficacy in Crohn’s disease197 CsA has no therapeutic value in Crohn’s disease. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines Monitoring therapy Measurement of blood pressure, full blood count, renal function and CsA concentration (aim for 100e200 ng/ml) are advisable at 0, 1 and 2 weeks, then monthly. Blood cholesterol and magnesium should be checked before starting therapy (see below).

Adverse effects of CsA Minor side effects occur in 31e51%, including tremor, paraesthesiae, malaise, headache, abnormal liver function, gingival hyperplasia and hirsutism. Major complications are reported in 0e17%, including renal impairment, infections and neurotoxicity. The risk of seizures is increased in patients with a low cholesterol (<3.0 mmol/l) or magnesium (<0.50 mmol/l). Oral therapy is an alternative in these circumstances. Prophylaxis against Pneumocystis carinii pneumonia is an individual decision dependent on nutritional state, concomitant immunosuppressive therapy and duration of therapy, but other opportunistic infections (eg, Aspergillus sp.) may be as common. Toxicity can be reduced by using lower doses (2 mg/kg/day intravenously),198 by oral microemulsion CsA,199 or by monotherapy without corticosteroids.194

Thiopurines after ciclosporin for induction of remission of severe ulcerative colitis CsA may be used as rescue therapy for steroid refractory acute severe ulcerative colitis (see section 5.3), but is best discontinued within 6 months because of nephrotoxicity. Consequently, immunosuppressives such as AZA or MP are often introduced while the patient is still on CsA and steroids are being tapered. The justiﬁcation of thiopurines in this setting (even in patients who may be 5-ASA naïve) is the high colectomy rate (36e69%) in the 12 months following introduction of CsA.200 The evidence that thiopurines reduce the risk of colectomy after an induction period with CsA is largely retrospective.201e203 Marion followed 29 patients for a median of 92 weeks and reported a 22% colectomy rate in those taking MP compared to 72% of those not taking MP. Similar results have been reported from Chicago, with 20% colectomy rate in those patients taking MP after CsA compared with 45% colectomy rate in those not taking MP.204 The Leuven group report experience with 142 patients, of who 118 (83%) had an initial response to CsA and avoided colectomy during initial hospitalisation.200 Sixty-four (54%) subsequently came to colectomy; the rate in those already on AZA compared to that in patients starting AZA concurrently with CsA was 59% vs. 31% (p<0.05). Life table analysis showed that 33% of patients came to colectomy at 1 year, with a probability rising to 88% at 7 years if CsA was used in patients already on AZA: The conclusion from this is that CsA has little role for patients who have failed AZA of an appropriate dose and duration.

Tacrolimus Tacrolimus is another calcineurin inhibitor often preferred in the transplant setting to CsA. Data from one placebo controlled trial and several series show that tacrolimus is effective in the treatment of steroid refractory thiopurine naïve ulcerative colitis.205 (Correction in Gut 2006;55:1684, regarding a dosage error in the abstract.)206e209 A dose is of 0.025 mg/kg tacrolimus twice a day should achieve trough levels of 10e15 ng/ml. Remission and colectomyfree survival are similar to oral and intravenous CsA. A direct comparison between tacrolimus and CsA has not been made.

4.4.7 Anti-TNF therapies http://www.bsg.org.uk/forum (accessed Oct 2010) There are presently two biological agents licensed for the treatment of IBD in the UK; both are monoclonal antibodies Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

against tumour necrosis factor a (anti-TNF). Inﬂiximab (IFX) is a chimeric anti-TNF antibody, consisting of 75% human IgG and 25% murine component that actively binds membrane-bound and soluble TNFa. IFX is given by intravenous infusion only. Adalimumab (ADA) is a humanised anti-TNF antibody, given by sub-cutaneous injection only. At the present time both agents are licensed for the treatment of inﬂammatory Crohn’s disease that has failed to respond to standard immunosuppression (ie, corticosteroids and thiopurine or methotrexate therapy). IFX is also licensed for ulcerative colitis and ﬁstulating Crohn’s disease.

Efficacy in Crohn’s disease Numerous large RCTs have documented the efﬁcacy of IFX and ADA for both inﬂammatory and ﬁstulating Crohn’s disease.

Efficacy for inflammatory Crohn’s disease210e212 A multi-centre, double-blind study in 108 patients with moderate-to-severe Crohn’s disease refractory to 5-ASA, corticosteroids and/or immunosuppressives, demonstrated an 81% response rate at 4 weeks after 5 mg/kg IFX compared with 17% given placebo. The duration of response varied, but 48% who had received 5 mg/kg still had a response at week 12. The ACCENT-1 study was the deﬁnitive re-treatment trial. Maintenance of remission in 335 responders to a single infusion of IFX 5 mg/kg for active Crohn’s disease (out of an initial 573) was examined. Patients were treated with placebo, 5 mg/kg or 10 mg/kg every 8 weeks until week 46. At week 30, remission rates were 21% in the placebo group compared to 39% in the 5 mg/kg group (p¼0.003) and 45% in the10 mg/kg group (p¼0.0002). ADA has demonstrated efﬁcacy in moderate to severely active luminal Crohn’s disease in both anti-TNF naïve patients and in those who failed IFX therapy. In CLASSIC-I, 30% of patients treated with ADA (160/80 mg or 80/40 mg) entered clinical remission versus 12% given placebo (p¼0.004).213 The remission rate at week 4 reached 35.5% in those patients loaded with 160/ 80 mg ADA. The GAIN study demonstrated efﬁcacy in patients who had previously failed IFX therapy (4 week remission rates of 21.4% vs. 7.2% placebo treated, p¼0.0006), albeit with lower remission rates than in CLASSIC-I.214 There was no statistical difference in GAIN between those failing IFX due to intolerance or as primary non-responders. The efﬁcacy of maintenance therapy has been conclusively demonstrated by the CLASSIC-II and CHARM studies.215 216

Efficacy for fistulating Crohn’s disease217

218

Present et al treated 94 patients with draining abdominal or perianal ﬁstulas of at least 3 months’ duration with IFX. 68% in the 5 mg/kg group and 56% in the 10 mg/kg group experienced a 50% reduction in the number of draining ﬁstulas at two or more consecutive visits (4 weeks apart) versus 26% given placebo (p¼0.002 and p¼0.02, respectively). However, the duration of this effect was in most cases limited to only 3 or 4 months. In a large re-treatment trial for ﬁstulating Crohn’s disease (ACCENT-II), 306 patients with actively draining abdominal or perianal ﬁstulae were treated with three induction infusions of IFX 5 mg/kg at week 0, 2 and 6. 195/306 (69%) responded and these were randomised to 5 mg/kg maintenance infusions or placebo every 8 weeks. Patients who lost response were switched from placebo to active treatment at 5 mg/kg, or the re-treatment dose increased from 5 to 10 mg/kg. At the end of the 12 month trial, 46% of the patients on active re-treatment had a ﬁstula response versus 23% on placebo (p¼0.001). Complete response (all ﬁstulae closed) was observed in 36% of patients on active treatment, compared to 19% on placebo (p¼0.009). Evidence for ﬁstula healing with ADA was provided in the CHARM study 583

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines

NICE guidance for the use of Adalimumab and Infliximab in the treatment of Crohn’s Disease 1.1 Infliximab and adalimumab, within their licensed indications, are recommended as treatment options for adults with severe active Crohn’s disease (see 1.6) whose disease has not responded to conventional therapy (including immunosuppressive and/or corticosteroid treatments), or who are intolerant of or have contraindications to conventional therapy. Infliximab or adalimumab should be given as a planned course of treatment until treatment failure (including the need for surgery), or until 12 months after the start of treatment, whichever is shorter. People should then have their disease reassessed (see 1.4) to determine whether ongoing treatment is still clinically appropriate. 1.2 Treatment as described in 1.1 should normally be started with the less expensive drug (taking into account drug administration costs, required dose and product price per dose). This may need to be varied for individual patients because of differences in the method of administration and treatment schedules. 1.3 Infliximab, within its licensed indication, is recommended as a treatment option for people with active fistulating Crohn’s disease whose disease has not responded to conventional therapy (including antibiotics, drainage and immunosuppressive treatments), or who are intolerant of or have contraindications to conventional therapy. Infliximab should be given as a planned course of treatment until treatment failure (including the need for surgery) or until 12 months after the start of treatment, whichever is shorter. People should then have their disease reassessed (see 1.4) to determine whether ongoing treatment is still clinically appropriate. 1.4 Treatment with infliximab or adalimumab (see 1.1 and 1.3) should only be continued if there is clear evidence of ongoing active disease as determined by clinical symptoms, biological markers and investigation, including endoscopy if necessary. Specialists should discuss the risks and benefits of continued treatment with patients and consider a trial withdrawal from treatment for all patients who are in stable clinical remission. People who continue treatment with infliximab or adalimumab should have their disease reassessed at least every 12 months to determine whether ongoing treatment is still clinically appropriate. People whose disease relapses after treatment is stopped should have the option to start treatment again. 1.5 Infliximab, within its licensed indication, is recommended for the treatment of people aged 6e17 years with severe active Crohn’s disease whose disease has not responded to conventional therapy (including corticosteroids, immunosuppressives and primary nutrition therapy), or who are intolerant of or have contraindications to conventional therapy. The need to continue treatment should be reviewed at least every 12 months. 1.6 For the purposes of this guidance, severe active Crohn’s disease is defined as very poor general health and one or more symptoms such as weight loss, fever, severe abdominal pain and usually frequent (3e4 or more) diarrhoeal stools daily. People with severe active Crohn’s disease may or may not develop new fistulae or have extra-intestinal manifestations of the disease. This clinical definition normally, but not exclusively, corresponds to a Crohn’s Disease Activity Index (CDAI) score of 300 or more, or a HarveyeBradshaw score of 8 to 9 or above. 1.7 When using the CDAI and HarveyeBradshaw index, healthcare professionals should take into account any physical, sensory or learning disabilities, or communication difficulties that could affect the scores and make any adjustments they consider appropriate. 1.8 Treatment with infliximab or adalimumab should only be started and reviewed by clinicians with experience of TNF inhibitors and of managing Crohn’s disease. where complete ﬁstula healing was noted in 33% compared with 13% given placebo at week 56 (p¼0.016) and thereafter maintained to 2 years in an open-label extension.215 In the UK, NICE has issued new guidance for clinicians on the use of IFX and ADA for the treatment of Crohn’s disease http:// guidance.nice.org.uk/TA187 (accessed Oct 2010) (summarised below) that offers a pragmatic approach to maintenance therapy. The IBD committee endorses this approach but recognises that practice in this area differs from some other nations. Several aspects, especially the optimal selection of patients and timing of stopping anti-TNF therapy lack a strong evidence base at this time and it is likely that practice will evolve in the near future.

Efficacy in ulcerative colitis At present, only IFX has been shown to be effective in ulcerative colitis. A recent Cochrane review of seven RCTs showed that IFX (three intravenous infusions at 0, 2 and 6 weeks) was more effective than placebo in inducing clinical remission (OR 3.22, 95% CI 2.18 to 4.76); inducing endoscopic remission (OR 1.88, 95% CI 1.54 to 2.28) and clinical response (OR 1.99, 95% CI 1.65 to 2.41) at 8 weeks.219 ACT1 was a 364 patient study in moderately active ulcerative colitis refractory to oral steroids and/or thiopurines, given placebo, IFX 5 mg/kg or 10 mg/kg, at 0, 2 and 6 weeks, then every 8 weeks for a year.220 The primary endpoint at week 8 (clinical response deﬁned as >30% and a three-point decrease in the Mayo activity index, with virtual 584

cessation of rectal bleeding) was reached by 37.2% (placebo), 69.4% (5 mg/kg) and 61.5% (10 mg/kg), p>0.001). The secondary endpoints of remission (14.9%, 38.8% and 32.0% respectively) and mucosal healing (33.9%, 62.0%, and 59.0%) were also achieved. Duration of effect was maintained through week 30 with rates of remission: 15.7%, 33.9% and 36.9%, p>0.001). In ACT2,220 an almost identical trial of a further 364 patients, where outpatients with moderately active ulcerative colitis refractory to 5-ASA could be included, the rates of response and remission at week 8 were 29.3%/5.7%, 64.5%/ 33.9% and 69.2%/27.5% in placebo, 5 mg/kg and 10 mg/kg respectively (p>0.001). Signiﬁcantly higher rates of mucosal healing were also demonstrated. Despite these positive ﬁndings, only 20.9% in the ACT1 study were in corticosteroid-free remission at 1 year follow-up. Colectomy rates at 54 weeks have recently been published and show an absolute risk reduction of 7% in the IFX treated group.221 There are no published RCTs examining the efﬁcacy of ADA in ulcerative colitis. However, case series suggest that ADA may have a role in treating mildemoderate ulcerative colitis patients who are intolerant, or have lost response to IFX.222e225

Efficacy for corticosteroid-refractory ulcerative colitis In a RCT involving 43 patients, Probert et al showed that IFX was not superior to placebo in inducing remission at week 6 (39% vs 30% remission rates in IFX and placebo groups respectively; p¼0.76).226 In this study, 77% of patients had received Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines corticosteroids for more than 14 days. In the ACT1/2 studies, a third of the patients were considered to be refractory to corticosteroids at the time of recruitment. Contrary to the earlier study, the response (not remission) rates were signiﬁcantly higher than placebo at week 8 (77% vs. 35%, p<0.001).220 NICE has not approved the use of IFX in subacute setting (deﬁned as outpatient basis) (http://guidance.nice.org.uk/TA140 (accessed Oct 2010)) based on the lack of cost effectiveness.

Efficacy for acute severe ulcerative colitis Jarnerot et al in a RCT involving 45 patients comparing a single infusion of IFX (5 mg/kg) and placebo showed signiﬁcantly lower colectomy rates within 90 days of therapy (primary endpoint; p¼0.017, OR 4.9).227 This study had a parallel design where either patients with a fulminant colitis index228 $8.0 on day 3 after institution of high dose intravenous corticosteroids; or a Seo index on day 5, 6, or 7 satisfying the criteria of a severe or moderately severe attack of ulcerative colitis unresponsive to standard therapy were included. In sub-group analysis, the patients in the latter group derived the most beneﬁt from IFX therapy (p¼0.009). Current NICE guidelines reccommend the use of ciclosporin as ﬁrst-line therapy in steroid refractory acute severe ulcerative colitis (and IFX use only if ciclosporin is contraindicated) based on health economic analyses and the paucity of data to support the use of IFX over ciclosporin (http://www.nice.org.uk/nicemedia/pdf/TA163Guidance.pdf (last accessed Oct 2010)). The CONSTRUCT UK trial will assess the comparative efﬁcacy of IFX versus ciclosporin in the acute severe ulcerative colitis setting, (http://www.crncc.nihr.ac.uk/ (last accessed Oct 2010)). The GETAID group are also due to report on a separate study (CYSIF) in this context.

Adverse effects of anti-TNF therapy Due to the nature of their effects on TNF, all anti-TNF therapies share a similar proﬁle of adverse events, including increased risk of infections from intracellular pathogens, most notably, TB, other opportunistic infections, autoimmunity, infusion reactions, and other more rare side-effects. This should be balanced with the potential curative option of surgery in ulcerative colitis.

Infections When considering this side effect of anti-TNF therapy, it is important to put it in context of corticosteroids and immunosuppressives, which also increase the risk of infectious complications. Toruner et al demonstrated that the respective use of corticosteroids, immunosuppressives or inﬂiximab conferred a threefold increased risk of developing an opportunistic infection that increased to 15-fold when two or more therapies are used in combination.229 In the Scottish severe ulcerative colitis cohort of 38 patients, inﬂiximab treatment as rescue therapy was associated with two serious adverse events: death secondary to pseudomonas pneumonia, and fungal septicaemia post-operatively.230 The recent Mayo Clinic experience shows that chronic ulcerative colitis patients treated with IFX before ileo-anal pouch anastamosis have substantially increased the odds of postoperative pouchrelated and infectious complications (odds ratio 2.7, CI 1.1 to 6.7). In this cohort, there was no mortality associated with IFX therapy.231 Further cohort studies have reported increased post-operative complications attributed to IFX,232 and increased post-operative complications attributed to corticosteroid use but not IFX.233 A subsequent meta-analysis has reported an increased risk of all post-operative complications with preGut 2011;60:571e607. doi:10.1136/gut.2010.224154

operative IFX use.234 Sub-group analysis was underpowered, but there was a trend towards increased infections. IFX therapy is associated with an increased risk of tuberculosis (TB).235 Pre-treatment screening for exposure to TB is important via a history, chest x-ray and tuberculin skin test if applicable. The British Thoracic Society has produced guidance for assessing risk of TB and managing disease in patients who are about to begin anti-TNF therapy (http://www.brit-thoracic.org.uk/Portals/0/ Clinical%20Information/Tuberculosis/Guidelines/antitnf.pdf).236 Where treatment for latent TB is needed 12 weeks therapy is recommended prior to initiation of anti-TNF therapy. Prophylactic treatment reduces the risk of TB by 70%.237 Diagnosis of latent TB in patients with IBD on immunosuppressives is difﬁcult as tuberculin skin testing has a high false negative rate. T cell interferon-gamma release assays are a more speciﬁc and probably a more sensitive test for diagnosis of M tuberculosis infection than the tuberculin skin testing in immunocompetent persons. Results are not affected by prior BCG vaccination. Data also suggest that results are unaffected by immunosuppression but are affected by current anti-TNF therapy.238 There is insufﬁcient evidence at present to recommend the use of interferon-gamma release assays, but NICE is examining their utility (http://guidance.nice.org.uk/ CG/Wave0/103 (last accessed Oct 2010)). Re-activation of chronic hepatitis B has been reported in patients treated with IFX.239 There are no data to suggest anti-TNF therapy has any effect on the course of chronic hepatitis C. Pre-treatment screening for exposure to hepatitis B is important; vaccination should be considered in the nonimmune high-risk patient. (see sections 6.0 and 7.8).

Antibody formation Antibodies to inﬂiximab (ATI) can trigger both acute infusion reactions and delayed serum-sickness-like reactions. Minor acute reactions usually respond to slowing the infusion rate or treatment with antihistamines, paracetamol and sometimes corticosteroids. Anti-histamines and steroids can be used as premedication to minimise anaphylactic reactions, which can be severe, especially after a prolonged drug holiday. Episodic therapy and consequent ‘drug holiday’ is associated with increased formation of ATIs, and should be avoided. In the ACCENT 1 study, the cumulative incidence of ATI was 30% through 72 weeks, signiﬁcantly higher than the 10% and 7% in the group of patients treated with systematic treatment with 5 or 10 mg/kg inﬂiximab infusion every 8 weeks. ATI formation is associated with increased incidence of infusion reactions and loss of response.240 Although ADA is a fully humanised antibody, it is also associated with the formation of antibodies to adalimumab (ATA) which have been shown to reduce efﬁcacy in rheumatoid arthritis241 and Crohn’s disease.242 There is emerging evidence linking low serum trough levels of IFX to lack of sustained response.243 244 Further research is required, but it appears serum IFX levels are inﬂuenced by ATIs and otherdprobably pharmacokineticdfactors. At this stage, it is not known what the target trough level should be.245 In the UK this issue is at present academic because there is no available commercial resource for measuring either trough levels or antibody levels. We think such a resource would be valuable.

Malignancy In a pooled analysis using results of placebo-controlled trials of IFX and ADA in patients with rheumatoid arthritis, the OR for malignancy (including basal and squamous cell cancers) was 3.3 (95% CI, 1.2 to 9.1).246 Of note, while 11 of the reported 35 585

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines malignancies in anti-TNF patients were lymphomas or leukaemias, 14 were solid organ cancers. The ﬁndings of this controversial meta-analysis were not observed in the TREAT registry137 and the Leuven dataset.247 In the combined analysis of 484 patients with ulcerative colitis who received IFX in the ACT trials there were four malignancies presenting in the trial period compared with one basal cell carcinoma in the 244 who received placebo. In an analysis of 38 patients from six English centres maintained on 8-weekly IFX infusions (median follow-up 15 months), there was one case of invasive breast carcinoma developing during treatment phase.248 The Mayo Clinic practice and Edinburgh series conﬁrmed the relatively rare occurrence of malignancy.249 250 However, some observations in these two latter studies serve to highlight possible groups of patients at higher risk of developing certain malignancies. In the Mayo Clinic series of 500 Crohn’s disease patients, two lung cancers, both thought ‘possibly related’ to IFX were reported in elderly smokers; and three lung cancers in 207 patients in the Edinburgh cohort. The Edinburgh group recently reported a case of non-small-cell lung cancer which resolved when anti-TNF therapy was discontinued.251 A recent 24-week trial of IFX in COPD was notable for the high malignancy rate in 157 IFX treated patients (nine malignancies including four lung cancers plus two additional lung cancers after study completion vs. 1/77 in placebo group).252 There is a recognised risk of non-Hodgkin’s lymphoma (NHL). Of recent particular concern is the recent reported cluster of the rare hepato-splenic T cell lymphoma in Crohn’s disease.253 This is now reported in both IFX and ADA therapy (17 cases in total) and also in ulcerative colitis (three cases).254 All except one were concomitantly treated with thiopurines; and the outcomes have been almost uniformly fatal. It is uncertain whether this is a combined cumulative effect of anti-TNF and thiopurine therapy. However, it is important to note that initial fears that there would be an epidemic of HSTL in anti-TNF treated patients have not been realised, with stable incidence rates. A recent meta-analysis examined the rate of NHL in adult Crohn’s disease in those who have received anti-TNF therapy and compared that to the rate in a population-based registry, and to the rate in those exposed to immunosuppressants alone.255 Anti-TNF therapy was associated with an increased risk of NHL when compared to the general population, but the risk remained small (6.1 per 10 000 patient-years). Anti-TNF therapy also led to an increased rate of NHL compared to those treated with immunosuppressants alone, although this did not reach signiﬁcance. One difﬁculty with interpreting these ﬁndings is that the majority of NHL cases patients had been exposed to immunosuppressants at some point. Thiopurines alone are associated with in increased risk of lymphoma162 163 and it is difﬁcult to establish the relative contribution of each drug.

Demyelination Reports of optic neuritis, seizure, and new onset or exacerbation of central nervous system demyelinating disorders, including multiple sclerosis, have been reported with the use of all antiTNFs. It is also noteworthy that in multiple sclerosis, a controlled trial of anti-TNF therapy (lenercept) resulted in a signiﬁcantly increased risk of exacerbation of disease.256 This may be relevant in high incident populations.

Congestive cardiac failure Anti-TNF agents are contraindicated for patients with class IIIeIV congestive heart failure due to evidence of increased risks of death from several clinical trials.

586

Unresolved issues concerning the use of anti-TNF therapy http://www.bsg.org.uk/forum (accessed Oct 2010) < In newly diagnosed patients with Crohn’s disease, which

patients should be offered early anti-TNF therapy? The topdown approach257 does not address this speciﬁc area and in fact, results in over-treatment in a considerable subset of patients with anti-TNF therapy (30%). This also has health economic implications. < It is unclear as to whether patients should be treated with solely anti-TNF monotherapy or with concomitant immunosuppression. The balance of argument pivots on the need to optimise anti-TNF therapy and the increased risk of complications associated with the latter. The recent SONIC study evaluated as primary end-point the efﬁcacy at 26 weeks of IFX monotherapy, AZA monotherapy and the two drugs combined in 508 patients with active Crohn’s disease who had not previously undergone immunosuppressive or biological therapy. On combination therapy, 56.8% were in steroid-free remission at week 26, compared with 44.4% of patients on IFX montherapy (p¼0.02), and 30.0% on AZA alone (p<0.001). A similar trend was reported at week 50. The study did not address longer-term efﬁcacy, safety, cost-issues, or withdrawal strategies, nor identify key prognostic factors for response to AZA alone, all critical issues in clinical practice.258 < In patients stably maintained in clinical remission (on antiTNF monotherapy or combination therapy), it is unclear how long therapy should continue or whether patients should wean off anti-TNF therapy or immunosuppressant therapy. In a small study of AZA withdrawal after 6 months of combination therapy (n¼80), continuation of immunosuppressants offered no clear beneﬁt over scheduled IFX monotherapy.259 However, a similar study (n¼48) reported IFX failure rates of 15% and 59% at 1 and 2 years respectively. Relapse was more likely in those with on-going inﬂammation.260 NICE recommends re-assessing the requirement for anti-TNF at 12 months, but there is no evidence to support any strategy. Emerging data in abstract form suggests that IFX may be discontinued and successfully re-introduced if patients relapse. < Could there be additional long term or unforeseen safety risks? Although a number of single-centred or tertiary antiTNF experiences have been reported,247 249 250 261 a framework to track and register any therapy-related complications that occur in clinical practice is necessary. This is particularly relevant in the face of the likely expansion in anti-TNF use, choice of anti-TNF agents and other newer biologics. The existing framework is not yet sufﬁciently robust to identify the more rare adverse events. For example, in pregnancy, a review of the FDA database reported 61 anomalies in 41 children exposed to anti-TNF agents in utero. Of these children, 24/41 (59%) had one or more congenital anomalies that are part of the vertebral abnormalities, anal atresia, cardiac defect, tracheoesophageal, renal, and limb abnormalities (VACTERL) association. There were 34 speciﬁc types of congenital anomalies in total, and 19 (56%) of those were part of the VACTERL spectrum.262 This study as been criticised for the lack of denominator and the fact that there was only one complete VACTERL anomaly, while other defects were mostly cardiac in origin. Further study is clearly warranted and caution advisable. A national biologics register is in progress. In general, the BSG committee favours a cautious measured approach to anti-TNF and other new biologics therapy in line with the primum non nocere principle.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 5.0 MANAGEMENT OF ULCERATIVE COLITIS http://www.bsg.org.uk/forum (accessed Oct 2010) Therapeutic decisions depend on disease activity and extent. Disease activity is best evaluated objectively using a clinical activity index (the Truelove and Witts263 or the Mayo Clinic disease activity index.264 Patients with severe disease require hospital admission, while those with mild/moderate disease can generally be managed as out patients. Disease extent can broadly be divided into distal and more extensive disease. However disease extent can vary (increase or decrease) with time in about 30% of patients. Topical management is appropriate for some patients with active disease. This is usually the case for those with proctitis and often the case if the disease extends into the sigmoid. For those with more extensive disease, oral or parenteral therapy is the mainstay of treatment, although patients may get additional beneﬁt from topical therapy.

5.1 Active left-sided or extensive ulcerative colitis7 82e86 103 265 Ulcerative colitis phenotype is determined by the maximal extent of inﬂammation observed at colonoscopy.23 ‘Left-sided’ ulcerative colitis (E2) is deﬁned as disease extending proximal to the recto-sigmoid junction up to the splenic ﬂexure; ‘extensive’ ulcerative colitis (E3) as extending proximal to the splenic ﬂexure (table 3). Disease activity should be conﬁrmed by sigmoidoscopy and infection excluded, although treatment need not wait for microbiological analysis.

Recommendations for the treatment of active (left-sided or extensive) ulcerative colitis:

< Oral mesalazine 2.4e4.8 g daily or balsalazide 6.75 g (deliv-

< <

<

<

ering 2.4 g mesalazine) daily are effective ﬁrst-line therapy for mildemoderately active disease (EL 1a, RG). Topical mesalazine combined with oral mesalazine >2 g/day is more effective than oral therapy alone for both left-sided (EL 1b, RG B) and extensive colitis (EL1b, RG A). Once daily dosing with mesalazine is at least as effective as twice or three times daily regimes. Prednisolone 20e40 mg daily is appropriate for those patients with moderately active disease, in whom mesalazine in appropriate dose and route has been unsuccessful (EL1b, RG C). Prednisolone should be reduced gradually according to severity and patient response, generally over 8 weeks. More rapid reduction is associated with early relapse. (For acute severe ulcerative colitis, see section 5.3).

5.2 Active proctitis7

83 89 129 266e269

Patient preference has a greater inﬂuence on management than for extensive colitis, in view of the option between topical or systemic therapy. Choice of topical formulation should be determined by the proximal extent of the inﬂammation (suppositories for disease to the recto-sigmoid junction, foam or liquid enemas for more proximal disease) along with patient preference, such as ease of insertion or retention of enemas. Proximal faecal loading is common in patients with distal colitis and may relate to a defect in colonic motility.269 It is a common cause of a patient not responding to apparently adequate therapy and is easily seen on a plain abdominal radiograph. Stool bulking agents are often not helpful: non-stimulant osmotic laxatives such as a polyethylene glycol (PEG)-based preparation are often helpful.

Recommendations for the treatment of active proctitis

< In mildemoderate disease, topical mesalazine 1e2 g daily (in

appropriate form for extent of disease) may be effective alone. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

<

<

<

<

Combination with oral mesalazine 2e4 g daily, or balsalazide 6.75 g daily, may be useful in resistant cases. (EL1b, RG B). Topical corticosteroids are less effective than topical mesalazine, and should be reserved as second-line therapy for patients who are unresponsive to topical mesalazine (EL 1b, RG B). Patients who have failed to improve on a combination of oral mesalazine with either topical mesalazine or topical corticosteroids should be treated with oral prednisolone 40 mg daily. Topical agents may be used as adjunctive therapy in this situation (EL1b, RG A). In the management of proximal faecal loading associated with distal colitis, non-stimulant osmotic laxatives such as a PEG-based preparation are often helpful (EL 3, RG C). Refractory proctitis should prompt exclusion of alternative pathology, consideration of drug compliance, change of formulation, associated irritable bowel, and further escalation of therapy.7

5.3 Acute severe ulcerative colitis194

195 198 227 270e272

Acute severe ulcerative colitis is a medical emergency. It is best deﬁned by the Truelove and Witts’ criteria263 (ulcerative colitis patients with $6 bloody stools/day and signs of systemic toxicity (HR>90, T>37.8, Hb#10.5 or ESR>30)). Patients should be admitted for intensive intravenous therapy. Intensive inpatient treatment with intravenous corticosteroids and early surgical intervention has reduced the UK mortality from acute severe ulcerative colitis to 2.9%.9 Seventy per cent of the deaths have signiﬁcant co-morbidity. However, the colectomy rate in acute severe ulcerative colitis (w30%) has not changed in 40 years.272 Acute ulcerative colitis is sometimes difﬁcult to distinguish from infective colitis; treatment should not be delayed until stool microbiology results are available. It may be appropriate to commence both corticosteroids and antibiotics. < The IBD service should have arrangements in place to admit known patients with IBD direct to the specialist gastroenterology ward or area. < Patients admitted with known or suspected IBD should discussed with and normally be transferred to the care of a consultant gastroenterologist/colorectal surgeon within 24 h of admission. < Where these facilities are not available (especially where there is no dedicated colorectal surgical service on site), patients should be transferred to appropriate centre for on-going joint medical-surgical management. Important steps in the initial management include: < Patients should be weighed and their nutritional needs assessed (IBD Standard A10). If the patient is malnourished nutritional support by the enteral route is associated with fewer complications than the parenteral route in acute coilitis.273 < Full blood count, urea and electrolytes, liver function tests, serum albumin, glucose and CRP, and haematinics (iron, B12, folate). Stool culture and C difﬁcile toxin assay. Microbiological testing for C difﬁcile. Toxin in addition to standard organisms in increasingly important. C difﬁcile infection has a higher prevalence in patients with IBD through unknown mechanisms, may not be conﬁned to the colon, and is associated with increased mortality. A minimum of four stool samples are required to detect 90% of cases.34 35 Cytomegalovirus (CMV) should be considered in severe or refractory colitis as reactivation is common in patients with IBD on immunosuppression and the 587

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines presentation can mimic ulcerative colitis or Crohn’s disease. CMV colitis is associated with a poor outcome and high colectomy rate.274e276 A combination of colonic histology and PCR for viral DNA conﬁrms the diagnosis rapidly. Immunosuppressants should be discontinued in favour of intravenous Gancyclovir for 2 weeks or the more expensive but equally effective oral agent Valgancyclovir. Further management is described in the ECCO consensus statement on prevention, diagnosis and management of opportunistic infections in IBD.277 < Intravenous ﬂuid and electrolyte replacement to correct and prevent dehydration or electrolyte imbalance, with blood transfusion to maintain a haemoglobin >10 g/dl. < Intravenous antibiotics only if infection is considered, or immediately prior to surgery. Withdrawal of anticholinergic, antidiarrhoeal agents, opioid drugs, which risk precipitating colonic dilatation. < Subcutaneous heparin to reduce the risk of thromboembolism.278 < Intravenous corticosteroids e Either hydrocortisone 100 mg four times a day or methylprednisolone 60 mg/day (EL1b, RGB). e Higher doses of steroids offer no greater beneﬁt, but lower doses are less effective. < If there is evidence of toxic megacolon (diameter >5.5 cm, or caecum >9 cm), the urgency with which surgery is undertaken after recognition of colonic dilatation depends on the condition of the patient: the greater the dilatation and the greater the degree of systemic toxicity, the sooner surgery should be undertaken, but signs may be masked by steroid therapy. In selected patients with mild dilatation, expectant management may be undertaken. < Any clinical, laboratory or radiological deterioration mandates immediate colectomy. < Flexible sigmoidoscopy and biopsy should be available within 72 h (ideally within 24 h) and a histological diagnosis within 5 days to conﬁrm diagnosis and exclude CMV. (IBD Standard A9) Daily monitoring: < Physical examination daily to evaluate abdominal tenderness and rebound tenderness. Joint medical and surgical management is appropriate (EL5, RG D). < Recording of vital signs four times daily and more often if deterioration noted. < Stool chart (to record number and character of bowel movements, including the presence or absence of blood and liquid versus solid stool). < Measurement of FBC, CRP, serum electrolytes, serum albumin, liver function tests and glucose every 24 h. < Consider need for daily AXR especially if there are signs of colonic distension and/or there is signiﬁcant deterioration in clinical condition or blood parameters. Further decision making: < A stool frequency of >8/day or CRP >45 mg/l at 3 days appears to predict the need for surgery in 85% of cases.270 Surgical review and input from specialist colorectal nurse or stoma therapist is appropriate at this stage. < Intravenous steroids are generally given for up to 5 days. There is no beneﬁt beyond 7e10 days.272 < Consideration of colectomy or rescue therapy with either intravenous ciclosporin (CsA) 2 mg/kg/day OR inﬂiximab (IFX) if there is no improvement by day 3 or there is subsequent deterioration (EL1b, RG B). NICE recommends CsA as ﬁrst-line (and IFX use only if CsA is contraindicated) based on health economic analyses and the paucity of data to support the use of IFX over CsA (http://www.nice.org.uk/ 588

nicemedia/pdf/TA163Guidance.pdf (accessed Oct 2010)). For patients already on immunosuppressive therapy such as AZA/MP at the time of presentation, surgery should be considered as the ﬁrst option (EL4, grade D). < Rescue with intravenous CsA: e 2 mg/kg/day is as effective as 4 mg/kg/day with decreased toxicity e Magnesium, cholesterol and creatinine should be measured within 48 h of starting CsA e Beware contraindications (Mg2+<0.5 mM or cholesterol <3.0 mM) and be vigilant for toxicity e Following induction of remission, oral CsA for 3e6 months is appropriate. (EL 1b, RG B). e Intravenous CsA alone may be as effective as methylprednisolone, but potential side effects mean that it is rarely an appropriate single ﬁrst line therapy (EL1b, RG C). < Rescue with IFX: e dose induction of 5 mg/kg (0, 2 and 6 weeks). The sideeffects of IFX, including therapy associated risk of mortality, should be discussed fully prior to its initiation (EL2, grade C). e IFX maintenance therapy in ulcerative colitis is not recommended because of the low corticosteroid-free remission rates after 1 year, and the limited data on subsequent need for colectomy (EL1b, grade C). e IFX should be given as a ‘bridge’ to longer term immunosuppressive therapy such as AZA/MP. < If no response to rescue therapy is seen within 4e7 days, colectomy is recommended (EL5, RG D). Speciﬁcally, we do not recommend CsA after IFX or vice versa (EL5, RG B). Other factors to consider: < The long-term follow-up of patients following an attack of acute severe ulcerative colitis reveals 50% of those who do not enter complete remission with steroids will require colectomy within 1 year.279 < Patients who avoid surgery should be considered for maintenance therapy with a thiopurine. (see section 4.2.4). < On discharge, oral steroids should be tapered over 8 weeks. Supplementation with calcium and vitamin D is recommended.280

5.4 Maintenance of remission7

82 84 90 93 103 105 157 168 265

Long-term maintenance therapy is generally recommended for all patients, especially those with left-sided or extensive disease, and those with proctitis who relapse more than once a year. Discontinuation of medication may be reasonable for those with distal disease who have been in remission for 2 years and are averse to such medication. The role of anti-TNF is discussed earlier and NICE does not approve use.

Recommendations for the maintenance of remission in ulcerative colitis:

< Patients with ulcerative colitis should normally receive

maintenance therapy with aminosalicylates, azathioprine, or mercaptopurine to reduce the risk of relapse. < Oral mesalazine 1.2e2.4 g daily or balsalazide 4.5 g daily should be considered as ﬁrst-line therapy (EL1b, RG A). < Topical mesalazine 1 g daily may be used in patients with distal disease with/without oral mesalazine, but patients are less likely to be compliant. (EL1b, RG B). < For patients on maintenance aminosalicylates, annual measurement of creatinine is sensible, although there is no evidence that monitoring is necessary or effective at preventing renal impairment. Aminosalicylates should be stopped if renal function deteriorates. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines < Long-term treatment with steroids is unacceptable. If steroids

cannot be withdrawn, surgery should be considered. < AZA 2e2.5 mg/kg/day, or MP 0.75e1.5 mg/kg/day are the ﬁrst line agents of choice for steroid-dependent disease.144 AZA is signiﬁcantly more effective than mesalazine at inducing clinical and endoscopic remission in the treatment of steroid dependent ulcerative colitis.143 (EL1b, RG A) These drugs should be considered in the following situations: e any patient who has a severe relapse or frequently relapsing disease e those who require two or more corticosteroid courses within a 12 month period e those whose disease relapses as the dose of steroid is reduced below 15 mg e relapse within 6 weeks of stopping corticosteroids e following ciclosporin (CsA) for induction of remission of severe ulcerative colitis (see section 4.4.6) < All patients should have measurement of thiopurine methyltransferase (TPMT) levels before starting thiopurines, mainly to avoid administration to a patient with no functional TPMT in whom thiopurine administration may be fatal (EL4, RG B) < For patients in remission, cessation may be considered after 4 years of full remission (EL2, RG C), but a small treatment beneﬁt persists even after 6 years (EL1b, RG B). Beneﬁts and risks of continuing thiopurines should be discussed with individual patients. < Methotrexate may be considered in the treatment of patients who do not respond to or are intolerant of thiopurines (EL4, RGC). Optimal duration of therapy is not established. < If ﬁrst-line immunosuppressive therapy fails, several factors (patient’s wishes, fecundity, patient age and extent of disease) need to be taken into account and a suitable therapeutic strategy to achieve steroid free remission discussed. In many cases this may necessitate surgery.

IBD service standards

< There must be local protocols for prescribing and monitoring

of thiopurines. Local practice should be audited (IBD Standard A6).

5.5 Surgery for ulcerative colitis http://www.bsg.org.uk/forum (accessed Oct 2010) Up to 30% of patients will ultimately require colectomy for ulcerative colitis.26 27 281 The decision to operate is best taken by the gastroenterologist and colorectal surgeon in conjunction with the patient.

5.5.1 indications

< < < < <

Disease not responding to intensive medical therapy Presence of dysplasia or carcinoma (see section 7.2) poorly controlled disease recurrent acute on chronic episodes of ulcerative colitis retained rectal stump following previous colectomy.

5.5.2 Operations There is debate over the efﬁcacy and safety of on-going medical therapy versus surgery in patients with acute severe colitis who fail initial high dose corticosteroids. Second-line medical therapy may reduce the need for immediate colectomy and yet many patients will relapse and require subsequent colectomy.279 Furthermore, second-line medical therapy carries with it a deﬁnable mortality risk200 and in comparison, in experienced Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

surgical hands, subtotal colectomy and ileostomy remains a safe alternative.282

Sub-total colectomy The operation of choice in patients with acute severe colitis failing to respond to intensive medical treatment is a subtotal colectomy, end ileostomy and preservation of a long rectal stump. The stump can be over-sewn and remain in the peritoneal cavity, sutured to the abdominal wall fascia beneath the skin or be delivered as a mucous ﬁstula. The choice of what to do depends upon the severity of disease in the rectum at the time of surgery.283 It is not recommended that patients undergo the ileoanal pouch procedure (IAPP) at this stage; they are often unwell, with low albumin and on high-dose steroids. A clinical colorectal nurse specialist in stoma therapy should perform preoperative counselling and marking of stoma sites.

Ileo-anal pouch procedure Patients requiring elective surgery for ulcerative colitis should be counselled regarding all surgical options. Surgery usually involves panproctocolectomy with permanent end ileostomy or IAPP. The choice between these two options is based upon patient preference and clinical criteria (ie, dysplasia/malignancy, sphincter injury or dysfunction). In appropriately selected cases it is difﬁcult to ﬁnd a difference in terms of quality of life between the two.284 There remain a number of controversies surrounding the IAPP with regard to technique: 1. one- versus two-stage procedures 2. hand-sewn or stapled pouch 3. pouch conﬁguration (W, S, J) 4. hand-sewn or stapled ileo-anal anastomoses The data to support one or any of these variations remains limited and it is difﬁcult to be precise regarding recommendations.285e290 Many of the choices rely on surgical judgement and surgical expertise. IAPP is a technically demanding procedure and carries with it a signiﬁcant morbidity rate and this morbidity relates to subsequent functional and quality of life outcomes.291e296 In addition, there is evidence that the success of IAPP in terms of this functional and quality of life outcome is related to some extent to the experience of the surgeon operating297 and the volume of the hospital practice.298 Furthermore, there is emerging evidence that units with greater experience of pouch surgery are better equipped to manage the complications and consequently preserve or improve outcome.299 It has therefore been suggested that surgical units undertaking IAPP should be performing at least ten cases per year as a minimum.7

5.5.3 complications and functional outcome Sub-total colectomy As a result of the severity of illness complications post surgery are signiﬁcant. Failure of healing and sepsis being common especially with patients on high does corticosteroids. There is evidence that delay in surgery as a result of prolonged ﬁrst or second line medical therapy may increase morbidity.300e302 This is further evidence for co-operative management of these patients by senior gastroenterologists and surgeons.

Ileo-anal pouch procedure While the functional outcomes following pouch procedures are favourable it remains a technically demanding procedure. Complication rates can be signiﬁcant and pouchitis remains a persistent and difﬁcult problem (section 5.6).303 Clinical outcomes after pouch surgery are variable in publishes series.295 304 The

589

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines latest data to emerge from the UK pouch registry suggest that the incidence of failure, deﬁned as excision or indeﬁnite diversion, was 7.7% following primary salvage. The median frequency of defaecation/24 h was ﬁve including one at night. Nocturnal seepage occurred in 8% at 1 year, rising to 15.4% at 20 years (p¼0.037). Urgency was experienced by 5.1% of patients at 1 year rising to 9.1% at 15 years (p¼0.022).305 Fecundity of young women may be reduced by 4050% following IAPP, probably as a result of pelvic surgery and subsequent pelvic adhesions.306e308 Appropriate informed consent, and an exploration of alternative medical or surgical options should be undertaken in women of childbearing potential before IAPP.

Recommendations for surgery in ulcerative colitis

< surgical units undertaking IAPP should be performing at least

10 cases per year as a minimum (EL5, RG D).

IBD Service Standards: (A7)

< expert pathological assessment before surgery is important < IBD surgery should be undertaken by colorectal surgeons in

a unit where the operations are performed regularly < pouch failure and salvage should be managed in a high-

volume specialist unit

5.6 Pouchitis http://www.bsg.org.uk/forum (accessed Oct 2010) Up to 50% of patients who undergo ileal pouch surgery for ulcerative colitis suffer from pouchitis.291 Symptoms include increased looseness and frequency of stool with or without bleeding. Urgency, tenesmus and pelvic discomfort in addition to fever and systemic upset may also occur.309 Diagnosis of pouchitis requires an appropriate clinical presentation in addition to endoscopic and histological conﬁrmation of inﬂammation.310 Conditions that mimic pouchitis (cufﬁtis, pelvic sepsis, prepouch ileitis, irritable pouch) should be considered.311e313

Treatment of pouchitis A number of trials exist to support the use of antibiotics and probiotics in the treatment and prevention of pouchitis.63 64 113 119 314e317 Other agents have been used in resistant pouchitis include budesonide,318 319 ciclosporin,320 short-chain fatty acids,321 and inﬂiximab.322 A Cochrane review has recently summarised the data.323

Recommendations for pouchitis

< The diagnosis of pouchitis should normally be made on

clinical and endoscopic and histological criteria (EL1a, RG A). < Metronidazole 400 mg three times a day (EL1a, RG A) or

ciproﬂoxacin 250 mg bd (EL1b, RG B) for 2 weeks is the ﬁrstline therapy of choice for pouchitis. < Long-term, low-dose metronidazole or ciproﬂoxacin are potentially effective for chronic pouchitis (RG B). < VSL#3 probiotic therapy may be used to treat and prevent pouchitis (EL2b, RG B). Its efﬁcacy is lost soon after stopping the treatment.

6.0 MANAGEMENT OF CROHN’S DISEASE The general principles are to consider the site (ileal, ileocolic, colonic, other), pattern (inﬂammatory, stricturing, ﬁstulating) and activity of the disease before treatment decisions are made in conjunction with the patient. Assessing the severity of Crohn’s disease can be more difﬁcult in comparison to ulcerative colitis. 590

An alternative explanation for symptoms other than active disease should be considered (such as infection, abscess formation, bacterial overgrowth, bile salt malabsorption, dysmotility, gall stones) and disease activity conﬁrmed before initiating new therapies. Routine blood tests including FBC, CRP and albumin provide an index of disease activity in some, but not all patients. The need for further imaging by endoscopy, barium radiology, CT scanning, MRI, EUA or capsule endoscopy should be considered in each individual patient. Smoking cessation should be strongly advised before discussion of any drug therapy (EL2b, RG C). NSAIDs are not recommended for pain relief in IBD. If NSAID administration is unavoidable, careful follow-up is suggested, as disease aggravation requires drug discontinuation (EL2b, RG B). Primary nutritional therapy should not be overlooked. Elemental or polymeric diets are less effective than corticosteroids, but may be used to induce remission in selected patients with active Crohn’s disease who have a contraindication to corticosteroid therapy, or where patients/physicans would prefer to avoid corticosteroids. (EL2b, RG C). Efﬁcacy is less in isolated colonic disease than small bowel disease and adherence a problem, especially in adults. Exclusive enteral nutrition (EEN) is appropriate as disease-modifying therapy for growth failure, and in adults may be used for those in whom corticosteroids are contraindicated or declined, or in patients who prefer EEN (EL2a, RG B). Before initiating biological therapy, steroids, or immunosuppression, infection needs to be rigorously excluded, and the potential role of surgery re-evaluated. Complex cases should be discussed at a MDT meeting. Careful discussion with each patient as to the likely beneﬁts/ risks of new therapies must be part of the decision-making process; these discussions should be documented in writing.

6.1 Active ileal, ileocolonic, or colonic Crohn’s disease2 3 32 75 84 145 146 173 177 178 210e218 http://www.bsg.org.uk/forum (accessed Oct 2010). In some, such as those with incidental disease detected at bowel cancer screening, therapy may not be required. In others, surgical review may be necessary at an early stage, often before initiating steroids, biological therapies or immunosuppressives. Early surgery may be preferable to medical therapy with many patients, and physicians. The risk of surgical complications is increased by delay to surgery, prolonged steroid usage, and malnutrition.324 325 Evidence for the use of antibiotics in shortterm therapy of colonic disease is available for metronidazole and ciproﬂoxacin. In practice, this form of therapy is limited for patients with refractory disease, or contra-indications to other therapies for which a stronger evidence basis exists. Inﬂiximab and adalimumab are efﬁcacious in inducing and maintaining remission in patients with active luminal Crohn’s disease and ﬁstulating disease affecting the perineum (see below). The drugs are relatively contra-indicated in the presence of ﬁbrostenotic disease, where surgical intervention is likely to be more appropriate. However, they may be preferable in extensive colitis or small bowel disease. Analysis of risks and beneﬁts of anti-TNF drugs needs be undertaken, balanced against other therapeutic alternatives and discussed with patients. In practice these discussions are best held in a multidisciplinary setting. The lack of evidence for anti-TNF therapy in a number of key areas needs to be addressed where possible by controlled trials: speciﬁcally, duration of therapy, long-term safety, and role of combination therapy. Current evidence and clinical experience now clearly favours scheduled maintenance therapy over episodic use of these agents. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines Recommendations for active ileal, ileocolonic, or colonic Crohn’s disease

< Initial treatment of active ileal or ileo-colonic Crohn’s disease

<

<

<

<

<

should be tailored to the severity of disease and must take the views of the patient into account. For patients with moderately active disease requiring treatment, oral corticosteroids such as prednisolone 2040 mg, or budesonide 9 mg daily is appropriate (EL1a). Prednisolone should be reduced gradually according to severity and patient response, generally over 8 weeks. More rapid reduction is associated with early relapse. If steroids are given, concomitant bone protection is recommended (see section 7). Budesonide 9 mg daily is appropriate for patients with isolated ileo-caecal disease with moderate disease activity. Although marginally less effective than prednisolone, its side-effect proﬁle is substantially better (EL1b). Failure to wean corticosteroids is common, and should be regarded as a treatment failure necessitating further intervention. In patients with severe active Crohn’s disease, or disease refractory to corticosteroids, anti-TNF therapy may be used in induction of remission, and in subsequent maintenance (see sections 4.4.7 and 6.2 for a detailed discussion of evidence, treatment strategies, and uncertainties that need to be addressed). AZA, MP and MTX are efﬁcacious in inducing remission, but each limited by time to response, side-effects, and uncertainties regarding drug withdrawal (see section 6.2)(EL1b, RG A).145 146 326 Refractory active Crohn’s disease remains an area for clinical trials of new therapies, supported by the National Institute of Health Research Portfolio, (http://www.crncc.nihr.ac.uk/ (accessed Oct 2010)).

IBD Service Standards

for those intolerant of, or who have failed to respond to thiopurines (EL2, RG B) once potential toxicity and other options, including surgery, have been discussed with the patient (see section 4.4.5 for details). < Anti-TNF therapy is effective in maintaining remission in Crohn’s disease (EL1a, RG A), although long-term data are lacking. It is best used as part of treatment strategy including immunomodulation once other options, including surgery, have been discussed with the patient. Treatment with ADA or IFX should only be started and reviewed by clinicians with experience of managing Crohn’s disease with anti-TNF therapy (http://guidance.nice.org.uk/TA187 (accessed Oct 2010)). Concurrent infection/sepsis should be excluded (see section 4.4.7) and treatment delayed until appropriate investigations (eg, cultures/imaging/examination under anaesthesia) and treatment (eg, antibiotics/surgical drainage) concluded.

Practical guidance in the use of anti-TNF therapies in induction and maintenance strategies

< For IFX, the dosing regimen is as follows:

< Access to a dietician and nutritional support should be

<

available to all patients with IBD (Standard A5). < Nutritional assessment should be performed on diagnosis and each hospital admission. Adolescents should have regular monitoring with height and weight centile charts and 6-monthly assessment of pubertal status (Standard A10).

<

6.2 Maintenance of remission

<

http://www.bsg.org.uk/forum (accessed Oct 2010). The majority of patients treated with steroids will not be in remission after 1 year133 and will therefore require maintenance therapy. Smoking cessation is probably the most important factor in maintaining remission and reducing the risk of relapse in Crohn’s disease.

Recommendations for maintenance of remission of Crohn’s disease

< AZA or MP should be considered as ﬁrst line treatment for

patients in the following situations (see section 4.4.4 for details): e any patient who has a severe relapse or frequently relapsing disease e those who require two or more corticosteroid courses within a 12 month period e those whose disease relapses as the dose of steroid is reduced below 15 mg e a relapse within 6 weeks of stopping corticosteroids < MTX is effective for patients whose active disease has responded to IM methotrexate (EL1b, RG A). It is appropriate

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

<

<

e A dose of 5 mg/kg IFX is used with loading doses at 0, 2 and 6 weeks: e If no evidence of initial response after two doses (primary non-responders), reconsider overall medical and surgical management of patient. Switch to ADA or dose intensiﬁcation to 10 mg/kg can be considered but with caution as data supporting these strategies are not strong (EL4, grade C). e If there is evidence of initial response, scheduled maintenance therapy will usually be appropriate. This is given initially at 8-weekly intervals (EL1b, RG B). Where response is lost, a valid initial strategy is to decrease infusion interval (initially to 6 weeks; no more frequent than 4-weekly) or to dose intensify by a single dose of 10 mg/kg or to switch to ADA (EL4, RG C). IFX should be used for ﬁstulating Crohn’s disease only after ensuring that all sepsis is actively draining; this requires appropriate cross sectional imaging (eg, MRI pelvis) and close collaboration with experienced colo-rectal surgeons (EL3, RG B). Pre-dosing with hydrocortisone is not usually required with the recommended scheduled maintenance IFX therapy. Initial infusions of IFX should be given over 2 h with close monitoring in a dedicated infusion facility, by trained personnel. Subsequent doses can be given over 1 h327 (EL4, RG D). If re-treatment with IFX is required after a signiﬁcant ‘drug holiday’ (>12 months) following initial IFX therapy, high vigilance is required for acute and chronic infusion reactions. Consider switching to alternative agent (ie, ADA) (EL5, RG D). For ADA the induction regimen can be 80 mg/40 mg sc on successive weeks, or 160 mg/80 mg215 (EL1b, grade B). The 80 mg/40 mg loading regimen is associated with a high requirement for subsequent does escalation.261 (EL4, grade C). The alternative of 160/80 mg may be more effective in patients who have lost response/intolerant to IFX.214 (EL2, grade C) e Maintenance therapy is 40 mg every other week e If response is lost, then escalate to 40 mg every week e If response is regained, it may then be possible to decrease dosing back to 40 mg every other week (EL5, RG D)

591

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines < For ADA therapy, the patient or relative/carer should be

<

<

<

< <

<

taught appropriate injection technique by an IBD nurse practitioner. Patients should be given clear advice about intercurrent illness (especially infection), when to delay treatment and who to contact for further advice. A medical history of demyelinating illness or optic neuritis is a relative contraindication for anti-TNF therapy (EL2b, RG C). In patients with a family history of demyelination, antiTNF therapy should be used with caution or avoided if possible. In this context, the risk of subsequent demyelinating episode is unclear. Expert neurological advice may be sought. The initiation of anti-TNF therapy during pregnancy should only be considered following full risk counselling with patient (and partner) particularly its unknown long-term effects. This should also be counterbalanced against risk of active disease in pregnancy and be applicable in patients already on maintenance anti-TNF therapy in Crohn’s disease (see section 7). Anti-TNF therapy should only be used with caution in older patients (>65 years old) with signiﬁcant smoking histories. If used in this situation, we suggest a CXR every 6e12 months (EL2b, RG C). We suggest particular caution in the use of anti-TNF therapy in patients with a medical history of malignancy. Anti-TNF therapy should be avoided in patients with congestive cardiac failure (EL2, RG B). In elderly patients or those with pre-existing ischaemic heart disease, the presence of cardiac failure should be screened (EL1b, RG B). NICE recommends that maintenance with anti-TNF therapy should continue until treatment failure (including the need for surgery) or until 12 months after the start of treatment, whichever is shorter: The disease should then be reassessed. Maintenance therapy should only then be continued if there is clear evidence of ongoing active disease as determined by clinical symptoms and investigation (http://guidance.nice. org.uk/TA187 (accessed Oct 2010)).

IBD service standards

< There must be local protocols for prescribing and monitoring

<

< <

< <

592

of thiopurines. Local practice should be audited (IBD Standard A6). Treatment with anti-TNF therapy should only be initiated by clinicians with expertise in their use for IBD (IBD Standard B1). Multi-disciplinary team meeting (IBD Standard A3). Where maintenance anti-TNF therapy is considered, it is recommended that the patient’s case is discussed at multidisciplinary IBD meeting where colorectal surgeons are also present. Of note, it should be considered at this stage whether or not surgery represents a more appropriate intervention for a particular patient. Counselling (IBD Standard C3) The risks and beneﬁts of treatment should be discussed with the patients and documented in the medical case records. In view of insufﬁcient evidence with respect to several key issues surrounding the long-term use of immunosuppressive therapy and anti-TNF therapy, limitations in knowledge need to be discussed with each patient. One mechanism of formalising such a discussion is the provision of formal consent. It is recommended that this conversation include the discussion of: e Efﬁcacy e Alternative treatment options, including surgery

e Risks of infection, infusion reactions, malignancy (including hepatosplenic T cell lymphoma), demyelination and drug-induced lupus eWritten information should be provided (IBD Standard D1)

6.3 Perianal Crohn’s disease4 http://www.bsg.org.uk/forum (accessed Oct 2010) The cumulative risk of developing a perianal ﬁstula is approximately 10% at 1 year, 15% at 5 years and 20% at 10 years.328 Fistulae were noted in 12% of patients with ileal disease, 15% with ileocolic disease, 41% with colonic and 92% with rectal involvement.329

Management principles329e335

The ﬁrst priority in the management of patients presenting with perianal sepsis/ﬁstula is to establish early and adequate surgical drainage. Clinicians should not be tempted to wait for the results of imaging investigations but should involve the appropriate surgical team to urgently examine the patient under general anaesthesia and to drain collections of pus. The next priority is to assess the extent of the disease. This can, in part, be achieved at the time of the surgical examination. In addition, MRI or ultrasound will help establish disease activity in the perineum. Luminal investigation is important to establish disease elsewhere, in particular in the rectum as this may inﬂuence longterm outcomes. At this stage patients with simple and low perianal ﬁstulae in the absence of rectal involvement can be treated with conventional surgical lay open. Patients with more complex disease or with rectal involvement need primary medical management and ‘adjuvant’ surgical intervention.

Recommendations for perianal Crohn’s disease Medical treatment4 145 146 210 217 218 257 331 336e341

< Metronidazole 400 mg three times a day and/or ciproﬂoxacin

500 mg bd are appropriate ﬁrst-line treatments for simple perianal ﬁstulae (EL4, RG D) (see section 4.2). < AZA 2e2.5 mg/kg/day or MP 0.75e1.5 mg/kg/day are potentially effective for simple perianal ﬁstulae or enterocutaneous ﬁstulae where distal obstruction and abscess have been excluded (EL4, RG D). < Anti-TNF therapy may be used in patients with severe perianal or enterocutaneous ﬁstulae or who are refractory to other treatment, and should be used as part of a pathway that includes immunosuppression and surgery.

Surgical treatment. Surgery should be used in conjunction with best medical therapy. Seton drainage can be a useful technique to provide symptom control and can be used as a prelude to medical treatment.342 343 Other surgical approaches such as advancement ﬂaps, and ﬁstula plugs may be appropriate for persistent or complex ﬁstulae in combination with medical treatment.344 There is insufﬁcient evidence to recommend other agents such as tacrolimus ointment and local inﬂiximab outside clinical trials or specialist centres. 6.4 Non-perianal fistulating Crohn’s disease4 http://www.bsg.org.uk/forum (accessed Oct 2010) This includes ﬁstulae communicating with other viscera (urinary bladder, vagina), loops of intestine (enteroenteral ﬁstulae), or the abdominal wall (enterocutaneous ﬁstulae). There is a notable lack of controlled data in this ﬁeld. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines Management of non-perianal fistulating Crohn’s disease There are no RCTs on the effect of medical treatment for nonperianal ﬁstulating Crohn’s disease, other than the subgroups of the ACCENT II trial. Less than 10% of the patients in the ACCENT II trial receiving inﬂiximab therapy had abdominal enterocutaneous ﬁstulae. For the 25 (of 282) patients with rectovaginal ﬁstulae in the ACCENT II trial, inﬂiximab was only modestly effective (45% closure at week 14).

Recommendations for entero-gynaecological fistulae

< Low anal-introital ﬁstula may be almost asymptomatic and

not need surgical treatment (EL5, RG D). < If the patient has a symptomatic ﬁstula, surgery is usually

necessary (including diverting ostomy) (EL5, RG D).

< Rectovaginal ﬁstulae failing conservative treatment should

have surgery with an advancement ﬂap and/or diverting ostomy if they are associated with unacceptable symptoms (EL5, RG D). < Intestinal small bowel or sigmoid-gynaecological ﬁstulae can usually be treated with resection of the diseased bowel segment (EL5, RG D).

Recommendations for enterovesical fistulae

< Surgery is the preferred approach for enterovesical ﬁstulae

(EL5, RG D). Only in high-risk patients (after multiple operations and\or severely shortened bowel), should medical therapy be the ﬁrst option (EL5, RG D).

Recommendations for enterocutaneous fistulae

< Post-surgical enterocutaneous ﬁstulae should initially be

treated with nutritional support and anatomical deﬁnition (EL5, RG D). Surgery after an interval is appropriate once nutrition is restored. < Primary enterocutaneous ﬁstulae can be treated medically but will generally require surgical management (by resecting the diseased bowel segment) (EL5, RG D).

6.5 Other sites The same general principles apply, although there are no randomised controlled trials.

Oral Crohn’s disease This is best managed in conjunction with a specialist in oral medicine. Topical steroids, topical tacrolimus, intra-lesional steroid injections, exclusion diets, enteral nutrition, and inﬂiximab may have a role in management but there are no randomised controlled trials.

Gastroduodenal disease

Disease in this area is associated with a poorer prognosis.345 Most clinicians would add a proton pump inhibitor to the conventional induction/maintenance therapies. Case reports describe a good outcome with anti-TNF therapy.346 Surgery is difﬁcult and may be complicated by ﬁstulation.

Post-surgical anastomotic strictures Endoscopic dilatation is an effective and safe treatment for short strictures and can delay requirement for further surgery.347 Injection of steroids into strictures may cause more harm.348

6.6 Surgery for Crohn’s disease http://www.bsg.org.uk/forum (accessed Oct 2010)

6.6.1 General principles During the lifetime of a patient with Crohn’s disease, surgery may be required in up to 75% of patients after 10 years of disease, Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

dependent on disease location.349 In addition, despite clear changes in the effective medical management of Crohn’s disease, evidence to date would suggest that there has been little change in the natural history of the disease and hence the need for surgery.349e351 Despite this the decision to operate remains difﬁcult. On the one hand complication rates and recurrence rates after surgery are relatively high.352 On the other hand, there is clear evidence that surgery provides good long-term disease control in many patients353 354 and that delay in operating may result in more advanced disease and hence more postoperative complications.324 325 The debate has been further fuelled by concerns over the effect of medical therapy on surgical complication rates. It seems clear that the preoperative use of steroids certainly does increase the subsequent surgical risk but current evidence suggests that immunosuppressive or biological agents do not.301 355 Laparoscopic surgery appears safe and feasible in Crohn’s disease and is emerging as the procedure of choice for ileocolic resections.356 The beneﬁts seem to be related to an improvement in early postoperative recovery, a reduction in wound complications and a cosmetic advantage. There is emerging evidence that there may be longer-term advantages in reducing subsequent adhesive complications and making subsequent resection in the event of recurrence possible laparoscopically as well.357e360 The management of colonic Crohn’s disease is perhaps more controversial. Segmental resection of isolated disease carries a higher recurrence rate but has all the functional beneﬁts of colonic preservation. It would appear that the beneﬁts are most obvious in right-sided disease and in patients with only one or two segments of isolated disease. Patients with predominantly left-sided disease or more than two areas involved do better with a subtotal colectomy and ileo-rectal anastomosis if the rectum is spared, or a panproctocolectomy and permanent ileostomy if the rectum is involved.361e363

6.6.2 Indications for surgery There are few randomised data to support decisions about surgery in Crohn’s disease and multidisciplinary meetings to discuss these issues are invaluable. Surgical intervention is governed by the extent of the disease, the response to medical treatment and the presence or absence of complications. Fibrostenotic and ﬁstulating intestinal disease with or without associated sepsis respond poorly to medical therapy; in the presence of a limited ileocolic distribution surgery is a good therapeutic option. In more extensive disease, preservation of bowel length is critically important. Limiting the resection to macroscopic disease and the use of strictureplasty have revolutionised surgery in this scenario.364e367

Recommendations for surgery in Crohn’s disease (IBD Service Standards: A7)

< Expert pathological assessment before surgery is important.

This may involve referral to a recognised expert in the differential diagnosis of IBD. < IBD surgery should be undertaken by colorectal surgeons (or their supervised trainees), who are core members of the IBD team in a unit where the operations are performed regularly.

6.6.3 Postoperative management: preventing postoperative recurrence http://www.bsg.org.uk/forum (accessed Oct 2010) Most patients with ileo-caecal Crohn’s disease who undergo surgical resection will develop recurrent disease in the neo-terminal ileum: endoscopic recurrence rates are 73% and 85% at 1 and 593

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 3 years, respectively.352 With no further medical therapy, clinical recurrence rates are about 2030% at 1 year, with a 10% increase per year in subsequent years.368 Observed recurrence rates are lower in patients who undergo resections from other sites.369 The most characterised risk factor for postoperative relapse is smoking. For patients who smoke, cessation signiﬁcantly reduces post-operative relapse.75 78 There has been some debate as to whether the technique of surgical anastomosis inﬂuences postoperative recurrence. Early reports suggested that a wide lumen stapled technique led to signiﬁcant reduction in recurrence as well as an increased time to recurrence.370 371 A meta-analysis suggested that these differences were at best minimal although the side-to-side anastomosis seemed to be associated with less complication and a lower leak rate than more traditional end-to-end techniques.372 Most recently a randomised multicentre trial has been completed to speciﬁcally address the issue of surgical anastomotic technique and recurrence. They concluded that there was no difference in recurrence rates between the two groups with the endoscopic recurrence rate being 42.5% in the end-to-end anastomosis group, compared with 37.9% in the side-to-side anastomosis group at a mean follow-up of 11.9 years.373

6.6.4 Medical therapy to prevent relapse At present, the evidence in favour of any speciﬁc medical intervention in the prevention of postoperative setting is weak, and open to criticism. While European guidelines suggest the use of mesalazine or thiopurines in this setting,4 a critical appraisal of the quality of evidence does not permit us to endorse this view, and we point to the need for randomised clinical trials to identify risk factors for recurrence, to deﬁne the efﬁcacy of currently available agents and the optimum timing of introducing and withdrawing these agents in this setting. Many clinicians will feel that it is necessary to rely on their clinical experience of treating active disease in deciding on therapy for the postoperative setting (in the absence of appropriate evidence) and prescribe thiopurines, metronidazole or mesalazine in perceived high-risk patients.374

Mesalazine The use of mesalazine therapy has been studied in postoperative prophylaxis in nine randomised clinical trials. Meta-analysis has been carried out but needs to be seen in the context of the heterogeneity of the designs of these trials; drug formulation, dosage, inclusion criteria, end-points, and duration of follow-up differed substantially and one trial was open-label rather than blinded. In the meta-analysis, the absolute risk difference between placebo, and mesalazine therapy is 10%, a ﬁnding of questionable clinical relevance.374

Thiopurines The two trials often quoted in support of thiopurine use in this context are now well-recognised as being substantially ﬂawed.374 In a trial with no pre-speciﬁed primary end-points, Hanauer et al375 used a ﬁxed dose of 50 mg MP with no doseeweight correction. The investigators demonstrated a statistically signiﬁcant beneﬁt over placebo at 2 years; however, by this stage 56% patients had withdrawn from the study, a factor that inevitably limits the interpretation of the data. Ardizzone376 compared AZA to mesalazine in postoperative prevention. There was no placebo group, and the two therapies showed no difference in the co-primary end-points of clinical and surgical recurrence at 2 years. Only a speculative post hoc sub594

analysis showed a beneﬁt for AZA in patients who had previously undergone a surgical resection. Problems in the design of this study are many and include the fact that the study was open label, not blinded, the lack of endoscopic conﬁrmation of clinical end-points, and ﬁnally the fact that the trial evaluated only conservative surgery such as strictureplasty or mini-resection allowing patients with residual active disease to be included. A meta-analysis of four of 15 potentially eligible studies including the two mentioned above looked at a total of 198 patients treated with AZA/MP and 235 control patients treated with mesalazine, placebo or metronidazole. It suggests that the effect of thiopurines in preventing postoperative recurrence is real but modest, averaging 813% at 1 year for clinical recurrence and 15% for endoscopic recurrence.377 We do not feel that any currently available evidence favours use of antibiotics, probiotics, prebiotics, corticosteroids, methotrexate or biologicals in this context; there are ongoing trials assessing the efﬁcacy of inﬂiximab in preventing postoperative recurrence. The ECCO consensus algorithm for management has been well received by many clinicians, with the objective of making decisions in high-risk patients on the basis of perceived risk factors as well as endoscopic appearances at 6 months. However, we point out that the strength of evidence to support risk factors other than smoking habit is weak, and the delay in instituting treatment until 6 months is such that this strategy should be interpreted as a treatment strategy for active disease, rather than primary prophylaxis.

6.6.5 Other considerations Parenteral vitamin B12 supplements There is some dispute about the length of resection required to produce deﬁciency of vitamin B12. Patients with less than 20 cm of ileal resection do not develop deﬁciency, while 52% of those with longer resection have abnormal Schilling tests.378 Resection of more than 60 cm almost invariably results in vitamin B12 deﬁciency.379 A practical approach is to treat with parenteral vitamin B12 all patients with more than 20 cm resected and measure serum B12 yearly in those with less than 20 cm resected.

Bile salt malabsorption Bile salt malabsorption occurs when normal active uptake from the ileum via the apical sodium dependant bile acid transporter is disrupted by ileal inﬂammation or resection, and results in watery diarrhoea. The degree of malabsorption depends on the length of ileal involvement or resection.380 Diagnosis of bile salt malabsorption can be made via SeHCAT scanning, although measurement of plasma lathosterol381 or serum 7-a-hydroxy cholestenone382 appears to give similar sensitivity in a simpler and less expensive test. Treatment is either with cholestyramine or a newer sequestrant such as colesevalam, which is more potent and better tolerated.

Recommendations for prevention of postoperative recurrence of Crohn’s disease

< Patients who smoke should be strongly advised to stop and

offered help to achieve this (EL2b, RG C).

7.0 ASSOCIATED ASPECTS OF INFLAMMATORY BOWEL DISEASE4 5 7.1 Management of pain and fatigue http://www.bsg.org.uk/forum (accessed Oct 2010) Abdominal pain is a common but under-researched feature of IBD. There are many potential mechanisms. These include acute and sub-acute obstruction in Crohn’s disease due to disease or Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines adhesions, serosal and mucosal inﬂammation, visceral hypersensitivity, secondary irritable bowel syndrome, proctalgia fugax, the likely impact of emotional factors on pain thresholds and visceral distension where there is dilation. Gallstones, renal calculi and chronic pancreatitis should be considered. In addition, arthritis, iritis and painful skin complications require analgesia in many patients. Most analgesics are relatively ineffective and have the potential to worsen underlying disease. Where possible, treatment is of the underlying cause (including corticosteroids and if appropriate, treatment of associated irritable bowel syndrome). Where non-speciﬁc pain relief is needed, an opioid that has less effect on motility, such as tramadol, may help. Patients with active and quiescent IBD often report symptoms of fatigue. Studies have demonstrated that fatigue measurement scores in patients with IBD are comparable to scores reported in cancer patients.383 As yet no identiﬁable cause has been found for ongoing fatigue in the absence of active disease; however, in patients complaining of this symptom it is important to exclude any clinical cause including anaemia and sub-therapeutic maintenance medication doses. In patients in whom no identiﬁable cause can be found, the symptom should not be ignored as it can have signiﬁcant consequences on the individuals’ quality of life, affecting work, school and social factors. More data is required on intervention strategies in this area.

7.2 Surveillance for colonic carcinoma28

93 384e387

http://www.bsg.org.uk/forum (accessed Oct 2010) Evidence relating to the increased incidence of colorectal carcinoma and the need for surveillance is reviewed in the ECCO consensus document.5 Recommendations reﬂect those published in the British Society of Gastroenterology Colonoscopic Surveillance Guidelines.388 Extensive colitis is deﬁned as ulcerative colitis extending proximal to the splenic ﬂexure (E3) or Crohn’s colitis affecting at least 50% of surface area of the colon (L4).23 (See tables 3 and 4). Patients with extensive colitis (ulcerative colitis or Crohn’s disease) can be risk stratiﬁed as follows: < Lower risk: 5-yearly colonoscopy e no endoscopic/histological active inﬂammation on the previous colonoscopy (histological chronic or quiescent changes acceptable) or e left-sided colitis (any grade of inﬂammation) or e Crohn’s disease colitis affecting <50% surface area of the colon (any grade of inﬂammation). < Intermediate risk: 3-yearly colonoscopy e mild endoscopic/histological active inﬂammation on the previous surveillance colonoscopy or e presence of post-inﬂammatory polyps or e family history of colorectal cancer in a ﬁrst-degree relative aged 50 years or over. < Higher risk: yearly colonoscopy e moderate or severe endoscopic/histological active inﬂammation on the previous surveillance colonoscopy or e stricture within past 5 years or e conﬁrmed dysplasia within past 5 years in a patient who declines surgery or e primary sclerosing cholangitis/post-orthotopic liver transplant for PSC or e family history of colorectal cancer in a ﬁrst-degree relative aged <50 years

Recommendations for the surveillance of colonic carcinoma

< The appropriateness of surveillance should be discussed with

patients who have ulcerative colitis or Crohn’s disease colitis Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

<

<

<

<

and a joint decision made on the balance of beneﬁt to the individual. The risk arising from ulcerative colitis and Crohn’s disease colitis is similar. Index (screening) colonoscopy is advised for all patients with ulcerative colitis or Crohn’s disease colitis at approximately 10 years after onset of symptoms to reassess disease extent (EL2, RG C). Surveillance colonoscopies should be performed, where possible, when the disease is in remission. However, a surveillance procedure should not be unduly delayed if remission cannot be achieved. Pancolonic dye spraying with targeted biopsy of abnormal areas is recommended. (EL2, RG A). If chromoendoscopy is not used, the strategy of random biopsy outlined in the 2002 surveillance guidelines should be followed. If a dysplastic polyp is detected within an area of inﬂammation and can be removed in its entirety, it is not necessary to recommend colectomy. Absence of dysplasia in surrounding tissue should be conﬁrmed.

Post-colectomy There is no clear evidence that pouch surveillance is beneﬁcial and thus it cannot be strongly recommended. However, if a clinician wishes to offer surveillance, the following surveillance policy would seem reasonable: < Higher-risk post-colectomy patients: consider yearly ﬂexible sigmoidoscopy of pouch/rectal mucosa in patients with: – previous rectal dysplasia or dysplasia or – colorectal cancer at the time of pouch surgery or – primary sclerosing cholangitis or – type C mucosa in the pouch (mucosa exhibiting permanent persistent atrophy and severe inﬂammation). < Lower-risk post-colectomy patients: consider 5-yearly ﬂexible sigmoidoscopy of pouch/rectal mucosa in patients with none of the risk factors above.

Recommendations for pouch surveillance

< Biopsies should be taken from pre-pouch ileum, the pouch-

anal anastomosis and the body of the pouch with four biopsies from each site. Pouch surveillance should be started early after pouch formation.

7.3 Management of pregnancy4 http://www.bsg.org.uk/forum (accessed Oct 2010) Inﬂammatory bowel disease is most commonly diagnosed in the third and fourth decades and therefore it is not unusual to care for females who are also pregnant. Approximately 25% of female patients conceive after the diagnosis of IBD. Although fertility in inactive IBD is unaffected, active disease diminishes this. This is further reduced by pelvic surgery or sepsis. Women with IBD are more likely to experience pre-term labour (<37 weeks duration) and low-birthweight babies (<2500 g).389 There is conﬂicting evidence regarding the frequency of foetal malformations. Many of the adverse outcomes in pregnancy are related to active disease rather than the medicines given to control this. Optimum disease control is necessary, ideally with remission prior to conception and active disease being treated aggressively to ensure best possible outcomes. Flares should be treated aggressively to prevent adverse outcomes (EL3a, RG B). Active disease is a risk for pre-term delivery and low birth weight (EL3a, RG B). Insufﬁcient data exist about maternal morbidity and foetal mortality at surgery. Combined care between gastroenterologists and obstetricians is mandatory. The medical management of pregnancy in IBD requires careful discussion with individual patients. There is a risk beneﬁt ratio of 595

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines each medicine taken that should be discussed ideally prior to conception. The Faculty of Sexual and Reproductive Health have produced clinical guidance that addresses fertility, pregnancy and contraceptive choice in patients with IBD, (http://www.ffprhc. org.uk/admin/uploads/810_CEUGuidanceIBD09.pdf (accessed Oct 2010)).

Medical treatment during pregnancy In active disease during pregnancy the Food and Drug Administration (FDA) pregnancy categories, ABCDX (see table) reﬂect a cautious approach. The drug description notice always emphasises risks and side effects.

Food and drug administration (FDA) categories A B C

Controlled studies show no risk No evidence of risk in humans Risk cannot be ruled out, animal studies showed adverse effects on fetus Positive evidence of risk in humans, risk/ benefit ratio should be considered Contraindicated

D X

The greatest risk to mother and fetus during pregnancy is active disease, and not the medication used to treat it. In general, pharmacological treatment for active disease during pregnancy is the same as for non-pregnant women.

Prescribing in pregnancy Considered safe

Probably safe

Contraindicated

Sulfasalazine (FDA B) Topical or oral mesalazine (FDA B) Corticosteroids (no rating)

Budesonide (FDA C) Thiopurines (FDA D)

Methotrexate (FDA X) Thalidomide (FDA X)

Infliximab (FDA B) Adalimumab Olsalazine Ciclosporin (FDA C) Quinolones (FDA C)

Tetracyclines (FDA D)

Tacrolimus (FDA C)

Diphenoxylate Metronidazole: avoid in first trimester Loperamide

Drugs used in IBD in pregnancy and breast feeding Aminosalicylates (FDA B) Sulfasalazine and the other 5-ASA drugs (up to 3 g/day) are safe in pregnancy and breast feeding. Sulfasalazine has the greatest clinical exposure. This leads to a reversible azospermia in men, there is a theoretical risk of neonatal haemolysis and it interferes with the absorption of folic acid. Folate supplementation is therefore recommended (2 mg/day) for women taking sulfasalazine. The safety of higher doses of 5-ASA is more uncertain. A single observational study identiﬁed low birthweight and preterm labour as consequences of treatment but the control group did not have IBD, a plausible explanation for this ﬁnding. Negligible amounts of 5-ASA are detected in breast milk as breast feeding is thought to be safe. Watery diarrhoea has been described in the infant but this usually ceases after withdrawing the medicine.

Antibiotics (FDA B&C) In the UK antibiotic use in IBD is most commonly used for exacerbations of perianal Crohn’s disease. The most commonly use agents are metronidazole and ciproﬂoxacin. Metronidazole is mutagenic in some strains of bacteria and carcinogenic in mice after long-term use but this has never been reported in humans. 596

Metronidazole is considered safe by most obstetricians after the ﬁrst trimester. Two studies examining the effect of the ﬂuoroquinolones on pregnancy have not identiﬁed any safety issues. Tetracyclines and sulfonamides should be avoided during pregnancy. Relatively large amounts of metronidazole are secreted in to breast milk and the manufacturer advises avoiding large single doses. No adverse events have been observed in humans.390

Corticosteroids (FDA C) Corticosteroids vary in their ability to cross the placenta; 88% of prednisolone is deactivated to a less active metabolite as it crosses the placenta resulting in low fetal blood concentrations. If administration is prolonged or repeated in pregnancy there is a risk of intra-uterine growth retardation but there is no evidence of this following short-term administration. Although the risk of cleft lip and palate, especially with ﬁrst trimester exposure is often cited391 the Committee on Safety of Medicines (CSM) in the UK concludes that there is no convincing evidence of such congenital abnormalities being associated with corticosteroids in humans. The balance of risks in mothers with active IBD unresponsive to other treatments would favour treatment with steroids. Corticosteroids are excreted in small amounts in breast milk but doses of prednisolone up to 40 mg daily are unlikely to cause systemic effects in the infant. Infants of mothers taking higher doses may theoretically have a degree of adrenal suppression but the beneﬁts of breast feeding are likely to outweigh the risks. A 4-h delay following oral exposure has been suggested to minimise exposure.

Budesonide (FDA C) No studies have been performed in patients with IBD but the theoretical risks are those in the section above. The use of inhaled budesonide would suggest that the drug is safe in pregnancy and breast feeding.

Thiopurines (FDA D) Although the manufacturers of AZA and MP advise avoiding in pregnancy there are considerable data, most of which come from the transplant and rheumatology literature, showing no alterations in fertility, pre-term delivery, or congenital defects. The high FDA rating is based on human reports of high abortion rates. There are, however, relatively little direct patient data available for IBD although that from the UK and US would suggest no increase in adverse outcomes. A recent Scandinavian population-based study of 900 children born to mothers with Crohn’s disease did suggest an increase risk of pre-term labour and congenital abnormalities in women exposed to both corticosteroids and AZA/MP.392 The total numbers of women on these medicines was relatively small and it was not possible to differentiate between the effect of the drugs and that of active disease. A single small study has suggested that when fathers used MP within three months of conception there was a higher incidence of pregnancy related complications. Therefore, although AZA and MP have FDA rating D, available data suggest that these drugs are safe and well tolerated during pregnancy. This is corroborated by the British National Formulary. Furthermore, although breast feeding is not advised, emerging data suggests there is very little exposure to the infant.393

Ciclosporin A (FDA C) Most available data regarding the use of ciclosporin in pregnancy originates from transplant and rheumatology literature. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines There is a higher rate of prematurity and low birthweight, but a high survival rate. A recent series of eight patients with acute severe ulcerative colitis did not identify any congenital malformations.394

Tacrolimus (FDA C) The transplant literature reports apparent safety. Preterm delivery is more common, but no excess congenital malformations, low birth weight, or neonatal complications have been found.

Methotrexate (FDA X) Methotrexate (MTX) is teratogenic and toxic and is contraindicated in pregnancy and breast feeding. If conception should accidentally occur, therapeutic abortion should be discussed, although not mandatory. MTX should be stopped immediately and high-dose folic acid replacement started. The intracellular metabolites of MTX, methotrexate polyglutamates, have a long half-life and take about 6 weeks to reach steady state or to completely wash out. Thus, women who wish to become pregnant should stop MTX for 36 months. The same applies to prospective fathers, to allow spermatogenesis return to normal.

Anti-TNF therapy (FDA B) (adapted from ECCO consensus4) Although there are reports of the use of inﬂiximab (IFX) and adalimumab (ADA) during pregnancy without apparent increased risk long-term data are not available.395 396 Anti-TNFa antibodies are species speciﬁc. Murine models have failed to show any teratogenicity or embryotoxicity. Case series from the rheumatology and gastroenterology literature have suggested good outcomes in relatively small numbers of patients. Postmarketing data from Centocor of more than 280 pregnancies, of which a third had IFX during the ﬁrst trimester showed that 75% had live births, 14% had a miscarriage, and 11% had therapeutic terminations.4 The long-term effects of anti-TNF therapy in utero on development, immune function and other biological programming are unknown. IFX and ADA are FDA Class B Risk Category in Pregnancy (similar to 5-ASA). Both are IgG1 antibodies that do not cross the placenta during the ﬁrst trimester of pregnancy. There is also limited data on anti-TNF excretion in breast milk. Maternal transfer of IFX has been reported in a patient using 10 mgs/kg body weight throughout pregnancy. The drug was seen to persist in the baby for 6 months but it is not known whether this induces antibody formation although this would seem unlikely. Placental transfer of IFX was not seen in three mothers using 5 mg/kg up until week 30.397 Infants exposed to anti-TNF therapy in utero are able to mount an appropriate immune response to vaccinations. Of some concern is a recent analysis of the FDA database documenting an increase in congenital abnormalities from the VACTERL spectrum262 in women taking etanercept or inﬂiximab (see section 4.4.7). Although this ﬁnding has not been conﬁrmed in further cohorts as yet, this demonstrates the need for cautious use in pregnancy. There needs to be a careful discussion of the risks and beneﬁts with individual patients. If treatment is continued it would seem reasonable to stop after the second trimester and to avoid dose escalation to prevent foetal transfer. Available data suggests that IFX cannot be detected in breast milk and thus breast feeding would on the available data appear safe. The implications of exposure to IFX on the newborn are unknown and a thorough discussion with patients is mandatory. Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Recommendations for drugs used in pregnancy

< Sulfasalazine should be stopped if there is suspected neonatal

haemolysis. < AZA should in general be continued during pregnancy, as the

risks to the fetus from disease activity appear to be greater than continued therapy. Babies born to mothers on AZA may be lighter than normal and the riskebeneﬁt should be discussed with patients. < Corticosteroids can be used for active disease, as the risks to the pregnancy from disease activity are greater than from continued therapy. < MTX is absolutely contra-indicated in pregnancy. < Physicians should exercise caution when considering the elective use of anti-TNF therapy in pregnant patients with IBD until further data become available regarding the frequency of congenital abnormalities and long-term outcomes. Conception should be discussed with women of childbearing age at the start of anti-TNF therapy. If treated patients present having become pregnant the treatment should be stopped after the second trimester.

Recommendations for the management of IBD in pregnancy

< A gastroenterologist and obstetrician should manage preg-

nant women with IBD jointly. < Maintaining adequate disease control during pregnancy is

<

<

<

<

<

essential for both maternal and fetal health. If planning pregnancy, patients should be counselled to conceive during remission and advised to continue their maintenance medication. Prior to conception patients should be well nourished and take folate supplements. Flexible sigmoidoscopy may be used safely where the information provided will signiﬁcantly alter management. The least extensive procedure possible should be employed. Patients with acute severe colitis or other life-threatening complications of disease should be managed as for the nonpregnant patient, including abdominal radiograph. The best interests of the fetus are served by optimal management of maternal IBD. The mode of delivery should be carefully considered. Patients with perianal Crohn’s disease or ileoanal pouch formation may best have a caesarean section to avoid the risk of damage to the anal sphincter. Absolute indications for surgery are unaltered by pregnancy and surgery should only be delayed where aggressive medical therapy may allow critical foetal maturation. Intestinal resection should be covered by a defunctioning stoma. Primary anastomosis is best avoided.

7.4 Management of extra-intestinal manifestations398 http://www.bsg.org.uk/forum (accessed Oct 2010) Extra-intestinal manifestations (EIMs) are found in both Crohn’s disease and ulcerative colitis, although they are commoner in Crohn’s disease (particularly colitis and ileocolitis). The commonest EIMs are musculoskeletal and mucocutaneous forms including axial and peripheral arthritis, acute ocular inﬂammation, erythema nodosum and pyoderma gangrenosum. The most signiﬁcant musculoskeletal manifestation is ankylosing spondylitis, which occurs in 15% of patients. This should be managed jointly with a rheumatologist, and may require biological therapy for the axial disease. In this case the choice of biological agent should be discussed between gastroenterologist and rheumatologist. Treatment for other EIMs consists of treating the underlying bowel disease, symptomatic relief and sometimes speciﬁc treatment of the EIM. Although 597

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines sulfasalazine has a higher incidence of side-effects compared to newer 5-ASA drugs, selected patients (such as those with a reactive arthropathy) may beneﬁt (RG D). It is important to try to avoid using NSAIDs for symptom relief, particularly in those patients with active gut disease. Peripheral arthritis is commonly associated with active disease, and normally responds to treatment of the bowel disease. For more persistent symptoms in the absence of active gut disease speciﬁc therapy may be required, including immune suppression and rarely biological therapy. Erythema nodosum is the commonest cutaneous manifestation, is usually associated with active disease, and responds well to treatment of the gut disease. Pyoderma gangrenosum is rare, but difﬁcult to treat and often requires treatment with calcineurin inhibitors or biological therapy. For acute ocular manifestations patients should be referred for ophthalmological assessment before starting therapy.

7.5 Primary sclerosing cholangitis399 http://www.bsg.org.uk/forum (accessed Oct 2010) Liver biochemistry may be abnormal in up to a third of patient with deﬁned IBD.400 Of this only 6% have a deﬁned liver disease of which primary sclerosing cholangitis (PSC) is the commonest (4.6%). Conversely, 70% of PSC patients have associated IBD. PSC is a rare but serious liver disease (incidence approximately 1:100 000 population/year). From diagnosis, the average survival varies from 12 to 17 years. High proportions of patients develop cirrhosis and require liver transplantation. There is a 5e15% lifetime risk of cholangiocarcinoma, which carries a poor prognosis. (See BSG guidelines on diagnosis and treatment of cholangiocarcinoma.)401 Several studies have indicated those patients with concomitant PSC are at a higher risk of colorectal neoplasia.385 402 The absolute cumulative risk of cancer or dysplasia in this subset of patients has been estimated to be 9% after 10 years, 31% after 20 years, and 50% after 25 years of colitis.403 Patients with PSC often have quiescent colitis and so it is difﬁcult estimating the exact onset of ulcerative colitis in this group. For the above reasons it is recommended such patients should have annual surveillance colonoscopy.384 The diagnosis of PSC is suggested by raised liver alkaline phosphatase, pANCA+, or changes of periductular ﬁbrosis on liver biopsy. The diagnosis requires stricturing and dilatation of the intra- and/or extrahepatic bile ducts on imaging. Magnetic resonance cholangiography (MRCP) avoids the risks of ERCP. Liver biopsy is necessary for diagnosis of small duct disease. Ursodeoxycholic acid (UDCA) improves liver biochemistry and at high dose may improve survival probability.404 However, a recent large RCT was stopped early due to excess adverse events in the group receiving high dose UDCA.405 Therefore high dose UDCA may be harmful.406 UDCA appears to reduce the risk of bowel cancer.407 408 Treatment of dominant strictures by ERCP and dilatation may be indicated and liver transplant is indicated for end stage liver disease.

7.6 Osteoporosis and osteomalacia Both osteoporosis and vitamin D deﬁciency (including compensated deﬁciency states with normal calcium and high parathyroid hormone) are common in IBD. The major risk factors for osteoporosis complicating IBD are age, steroid use and disease activity. The reader is referred to the 2007 BSG guidelines for osteoporosis in IBD and coeliac disease for a comprehensive review (http://www.bsg.org.uk/pdf_word_docs/ ost_coe_ibd.pdf (last accessed Oct 2010)) along with the guidelines of the Royal College of Physicians (http://bookshop. 598

rcplondon.ac.uk/contents/pub89-a953a6c0-06c0-46d8-b79ae951536d9070.pdf (last accessed Oct 2010)).

Recommendations for osteoporosis and osteomalacia:

< Supplementation of calcium and vitamin D is recommended

when systemic steroid use is necessary (EL3, RG C). of bisphosphonates with steroids is recommended for patients aged over 65 years or with known osteoporosis/osteopenia. Unless advised on other grounds, the bisphosphonate should only be given while the patient is on steroids (EL4, RG C).

< Co-administration

7.7 Anaemia http://www.bsg.org.uk/forum (accessed Oct 2010) Anaemia is a common complication of IBD.409 Comprehensive guidelines have recently been published from an expert working group.410 Iron deﬁciency and anaemia of chronic disease are the commonest causes of anaemia in IBD, though folate and vitamin B12 deﬁciency occur. Drug-induced anaemia secondary to AZA, MP or sulfasalazine also occurs. Other coexisting causes of anaemia such as menorrhagia and coeliac disease should be sought by careful history taking and use of coeliac serological testing. In older patients and those with a family history of cancer, investigation should exclude bowel cancer as a cause.

Screening for anaemia Patients with IBD should have at least annual haemoglobin check. The ferritin, transferrin saturation and CRP should be checked in anaemic patients or those with low MCV. The CRP is important to interpret the ferritin level as ferritin can be elevated in an acute phase reaction. Levels of ferritin less than 100 mg/l are suggestive of iron deﬁciency.411 Those patients with small bowel disease at risk of folate or B12 malabsorption or with a macrocytosis should have levels of B12 and folate checked.

Treatment of iron deficiency Long-term prevention of anaemia by treatment of underlying IBD is primary412 but iron replacement is also needed and improves quality of life. Treatment may be with oral iron, (eg, ferrous sulfate 200 mg bd or another preparation with equal amounts of elemental iron, c. 130 mg/day), but this may not be tolerated well and may exacerbate IBD symptoms measured by activity scores. In patients with poor tolerance to oral iron, intravenous replacement is preferred. Iron sucrose (Venofer), ferric carboxy-maltose (Ferinject) do not have the magnitude of risk of anaphylaxis of iron dextran and are generally well tolerated and usually effective.413e415

Other anaemias B12 and folate deﬁciency may occur in Crohn’s disease and replacement with oral folate and IM B12 is appropriate. The need for B12 replacement therapy should be anticipated in patients who have had ileal resections (see section 6.6.5). Monitoring or early replacement should be instituted. Thiopurines also cause macrocytosis and anaemia. If vitamin levels and iron are normal then drug-induced anaemia should be considered. Referral for haematology opinion and bone marrow examination may be necessary along with the considered withdrawal of any implicated drug treatment.

Non-responsive anaemia In patients with IBD and severe anaemia that is non responsive to iron therapy there is good evidence to show that erythropoietin analogue therapies will produce a response in 70100% Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines of patients.410 416 Cost, however, is a limiting factor and recent UK NICE guidelines for anaemia induced by cancer therapy have limited the use in the context of cancer chemotherapy to patients unable to have transfusions and with Hb of less than 8 g/dl. The cost in 2008 was from £188 to £234 per week. (NICE technology appraisal guidance 142: Epoetin alfa, epoetin b and darbepoetin alfa for cancer treatment-induced anaemia; www. nice.org.uk/TA142 (last accessed Oct 2010)).

lence in IBD 27%419 and 12 month point prevalence 15%.420 Anxiety is also more common in patients with IBD than in controls.421 Mood disorders in patients with IBD are at least in part a consequence of the IBD itself422 and its medical treatment (eg, corticosteroid therapy); surgery, including speciﬁcally colectomy and stoma formation also have psychosocial implications as do awareness of the risk of cancer and cancer surveillance.

7.8 Vaccinations

7.9.2 Psychological stress as a trigger for disease or relapse?

http://www.bsg.org.uk/forum (accessed Oct 2010) Patients with IBD may be at risk for infections due to underlying disease, malnutrition, surgery, or immunosuppressive therapy. The available data on the incidence and prevention of opportunistic infections in IBD has recently been published and summarised in a leading article in Gut.277 417

7.8.1 Infection and immunisation history A vaccine and infection history is best taken at baseline when a patient is diagnosed with IBD, including TB exposure, chickenpox history and risk of hepatitis B. Varicella zoster serology is best checked if there is no history of infection. We recommend checking hepatitis B serology in high-risk patients and prior to anti-TNF therapy (see section 4.4.7). If patients are sero-positive for hepatitis B, please refer to the ECCO consensus document on prevention, diagnosis and management of opportunistic infections in IBD for guidance.277

7.8.2 Recommended vaccinations

< Inﬂuenza, pneumococcal and HPV (females) vaccination is

generally recommended for immunosuppressed adults and is best considered for all patients with IBD, given the frequent need for steroid and immunosuppressive therapy. Booster vaccinations are appropriate for inﬂuenza (annually) and pneumococcus (after 3 years).418 < Hep B vaccinations should be considered prior to immunosuppressive or anti-TNF monoclonal antibody therapy in the non immune high-risk patient. < Live vaccines should be avoided in patients on immunosuppression or steroids (MMR, oral polio, yellow fever, live typhoid, varicella, BCG). < Varicella vaccination before treatment with steroids or immunosuppressants is now a possibility and has been recommended in Europe and the USA in the non-immune.

7.8.3 Post-exposure prophylaxis Post-exposure prophylaxis of varicella and measles exposed nonimmune individuals on high-dose steroid or immunosuppression is appropriate with immune globulin (varicella zoster immunoglobulin or human normal immunoglobulin).418 Aciclovir prophylaxis may also be used for varicella.

Recommendations for vaccinations

< A vaccination and infection history should be taken in all

patients with IBD (EL5, RG D). < Primary and booster vaccination for inﬂuenza and pneumo-

coccus should be offered to immunosuppressed patients with IBD.

7.9 Psychological aspects http://www.bsg.org.uk/forum (accessed Oct 2010).

7.9.1 Incidence and prevalence of mood disorders in IBD The incidence of depressive illness is higher in IBD cohorts than control populations with a reported OR of 2.2 (lifetime prevaGut 2011;60:571e607. doi:10.1136/gut.2010.224154

Human and animal studies have revealed psycho-neuroimmunological mechanisms whereby stress could inﬂuence the course of IBD.423 Stress and adverse life events do not appear to trigger the onset of Crohn’s disease or ulcerative colitis, but most reports indicate that they may be involved in triggering relapse of IBD.424 425 Furthermore, behaviour limiting exposure to stressful situations is associated with reduced symptomatic relapse, at least in Crohn’s disease.426

7.9.3 Effectiveness of psychological support in IBD The effectiveness of psychological interventions has been reviewed by ECCO for both ulcerative colitis and Crohn’s disease.4 5 Evidence indicates that psychosocial support is useful, particularly in adolescents.There is no deﬁnitive evidence that psychological interventions improve the course of IBD itself but they do usually improve patients’ quality of life and wellbeing.427 428 In general, psychological and psychiatric support should be made available where psychological concerns are present. (The psychological care of medical patients: A practical guide. Royal College of Physicians Royal College of Psychiatrists Report of a joint working party of the Royal College of Physicians and the Royal College of Psychiatrists Second edition 2003 http://www.rcplondon.ac.uk/pubs/wp/wp_pcomp.pdf (last accessed Oct 2010); IBD Service standards http://www. ibdstandards.org.uk. IBD nurses may play an important role in this regard either with formal training or informally.

IBD service standard

< Psychological support should be available to patients with

IBD (IBD Standard A2).

7.10 Inflammatory bowel disease in children http://www.bsg.org.uk/forum (accessed Oct 2010) Guidelines produced by the British Society of Paediatric Gastroenterology, Hepatology and Nutrition can be accessed from the society website: http://www.bspghan.org.uk/ (last accessed Oct 2010). Information for parents and patients is available from the Crohn’s in Childhood Research Association (CICRA) www. cicra.org (accessed Oct 2010).

7.11 Management of adolescents (transitional care) http://www.bsg.org.uk/forum (accessed Oct 2010) Guidelines for transition of patients with IBD were published by CICRA and NACC in 2008 www.nacc.org.uk (accessed Oct 2010). There are three separate documents for the professionals, parents and the patients.

Definition Transition is the planned move of adolescents and young adults with IBD from child-centred to adult-orientated healthcare and is a process, not a single event. Transfer is the successful handover of care to adult services.429 The National Service Framework for Children, Young People and Maternity Services’ guidance deﬁnes transition as: 599

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines ‘A purposeful, planned process that addresses the medical, psychosocial and educational/vocational needs of adolescents and young adults as they move from child-centred to adult-oriented healthcare systems.’ (Department of Health Transition: Getting it right for young people. National Service Framework for Children, Young People and Maternity Services, 2006) http://www.dh.gov. uk/en/Publicationsandstatistics/Publications/PublicationsPolicy AndGuidance/Browsable/DH_4132944 (accessed Oct 2010).

The principles that inform the guidelines Young people with IBD have a right to a managed transition process when moving from paediatric to adult care. They should be continuously prepared for transition throughout their teens to ensure they are ready for formal transfer of care to the adult services. Transition should not compromise the young person’s current care or their treatment options. It begins in paediatric services but adult services bear responsibility for its successful completion. Many factors are important in timing the transfer from child to adult care. The young person (and parents if the young person wishes) should be involved or represented in planning their transition. Young people need well-developed social, interpersonal and emotional skills to successfully enter the world of adult healthcare. Transition works best when it is coordinated and overseen by a nominated key worker or coordinator430 (Department of Health. Getting the right start: National Service Framework for Children. Standard for Hospital Services.2003 paras 4.58e4.62, http://www.dh.gov.uk/en/Publicationsandstatistics/ Publications/PublicationsPolicyAndGuidance/DH_4006182 (accessed Oct 2010)). Care plans including details of investigations and treatments tried successfully or otherwise must be drawn up. Multidisciplinary teams of paediatric and adult health professionals working together should provide transitional care. (Department of Health. National Service Framework for Children, Young People and Maternity Services. Core Standards. 2004) (see link above).

IBD service standard: (Standard A12)

Ho has received support to attend academic meetings from Abbott, Schering-Plough, Shire, and Proctor and Gamble. He has also received honoraria from Abbott, Shire, Otsuka Pharmaceuticals and Astra Zeneca for presentations at academic meetings. Dr Charlie Lees has acted as a consultant to Abbott and Dr Falk, and received honoraria from Shire, Procter and Gamble, Ferring, and Schering-Plough for presentations at academic meetings. Mr Alastair Windsor has received support to attend academic meetings from Proctor and Gamble. He has sat on advisory boards for Procter and Gamble and Ferring. He has also received honoraria from Abbott, Proctor and Gamble, and Ferring for presentations at academic meetings. Dr Stuart Bloom has acted as a consultant for and sat on advisory boards for Abbott, Shering-Plough, Shire, Ferring, Proctor and Gamble. He has received honoraria from Shering-Plough for chairing symposia. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

< There must be a deﬁned policy and protocol for transitional

care. < A named co-ordinator should be responsible for the prepara-

tion and oversight of transition (eg, IBD nurse specialist).

14. 15. 16.

Acknowledgements The authors would like to thank Belinda Theis for assisting with literature searches, editorial meetings and proof reading. We thank Simon Travis for making the revised set of ECCO consensus statements available ahead of their publication. Competing interests Dr Craig Mowat has received support to attend academic meetings from Abbott, Shire, Ferring and UCB, and has sat on advisory boards for Abbott, Shire, and Ferring. He has received honoraria from Schering-Plough, Shire and Ferring for presentations at academic meetings. Dr A.T. Cole has received educational grants from Abbott, Scherring-Plough, Ferring and Dr Falk Ag in support of conference attendance and to support educational meetings and has been on advisory boards for Schering-Plough and Ferring. Dr Ian Arnott has sat on advisory boards for Abbott and Schering-Plough. He has received support to attend academic meetings from Abbott, Shire, and Proctor and Gamble. He has also received honoraria from Abbott, Schering-Plough, Shire and Ferring for presentations at academic meetings. Professor Jack Satsangi has acted as a consultant to Schering-Plough, Abbott, UCB, Shire and Ferring, and has on-going research support from Novartis and Genentech. Dr Matt Rutter has received support to attend academic meetings from Abbott, Shire and Proctor and Gamble. He has also received honoraria from Shire, Proctor and Gamble, and Ferring for presentations at academic meetings. Dr Tim Orchard has sat on advisory boards for, and has received support to attend academic meetings from Abbott, Schering-Plough, Dr Falk Pharma, Ferring, Procter and Gamble, Shire and UCB. He has received honoraria from Ferring, Abbott, Schering-Plough, Procter and Gamble, and Shire. He has on-going research support form Johnson and Johnson. Dr Gwo-Tzer 600

17. 18. 19. 20. 21. 22. 23.

24. 25.

Carter MJ, Lobo AJ, Travis SP. Guidelines for the management of inflammatory bowel disease in adults. Gut 2004;(53 Suppl 5):V1e16. Stange EF, Travis SP, Vermeire S, et al. European evidence based consensus on the diagnosis and management of Crohn’s disease: definitions and diagnosis. Gut 2006;(55 Suppl 1):i1e15. Travis SP, Stange EF, Lemann M, et al. European evidence based consensus on the diagnosis and management of Crohn’s disease: current management. Gut 2006; (55 Suppl 1):i16e35. Caprilli R, Gassull MA, Escher JC, et al. European evidence based consensus on the diagnosis and management of Crohn’s disease: special situations. Gut 2006;55: i36e58. Biancone L, Michetti P, Travis S, et al. European evidence-based Consensus on the management of ulcerative colitis: special situations. J Crohns Colitis 2008;2:63e92. Stange EF, Travis SPL, Vermeire S, et al. European evidence-based consensus on the diagnosis and management of ulcerative colitis: definitions and diagnosis. J Crohns Colitis 2008;2:1e23. Travis SPL, Stange EF, Lemann M, et al. European evidence-based consensus on the management of ulcerative colitis: current management. J Crohns Colitis 2008;2:24e62. UK IBD Audit 2006: National Results for the Organisation & Process of IBD Care in the UK. Royal College of Physicians of London, 2006. Arnott Idr LK, Down C, Rhodes J, et al. Outcome of acute severe ulcerative colitis: data from the UK National IBD Audit. Gut 2009;58:A33. Hayward RS, Wilson MC, Tunis SR, et al. Users’ guides to the medical literature. VIII. How to use clinical practice guidelines. A. Are the recommendations valid? The Evidence-Based Medicine Working Group. JAMA 1995;274:570e4. Quality Care Service Standards for the Healthcare of People who have Inflammatory Bowel Disease (ibd): Oyster Healthcare Communications Ltd, 2009. Rubin GP, Hungin AP, Kelly PJ, et al. Inflammatory bowel disease: epidemiology and management in an English general practice population. Aliment Pharmacol Ther 2000;14:1553e9. Thomas GA, Millar-Jones D, Rhodes J, et al. Incidence of Crohn’s disease in Cardiff over 60 years: 1986e1990 an update. Eur J Gastroenterol Hepatol 1995;7:401e5. Logan RF. Inflammatory bowel disease incidence: up, down or unchanged? Gut 1998;42:309e11. Armitage E, Drummond H, Ghosh S, et al. Incidence of juvenile-onset Crohn’s disease in Scotland. Lancet 1999;353:1496e7. Gunesh S, Thomas GA, Williams GT, et al. The incidence of Crohn’s disease in Cardiff over the last 75 years: an update for 1996e2005. Aliment Pharmacol Ther 2008;27:211e19. Munkholm P, Langholz E, Davidsen M, et al. Disease activity courses in a regional cohort of Crohn’s disease patients. Scand J Gastroenterol 1995;30:699e706. Langholz E, Munkholm P, Davidsen M, et al. Course of ulcerative colitis: analysis of changes in disease activity over years. Gastroenterology 1994;107:3e11. Munkholm P, Langholz E, Davidsen M, et al. Intestinal cancer risk and mortality in patients with Crohn’s disease. Gastroenterology 1993;105:1716e23. Lennard-Jones JE, Shivananda S. Clinical uniformity of inflammatory bowel disease a presentation and during the first year of disease in the north and south of Europe. EC-IBD Study Group. Eur J Gastroenterol Hepatol 1997;9:353e9. Winther KV, Jess T, Langholz E, et al. Survival and cause-specific mortality in ulcerative colitis: follow-up of a population-based cohort in Copenhagen County. Gastroenterology 2003;125:1576e82. Luces C, Bodger K. Economic burden of inflammatory bowel disease: a UK perspective. Expert Rev Pharmacoecon Outcomes Res 2006;6:471e82. Silverberg MS, Satsangi J, Ahmad T, et al. Toward an integrated clinical, molecular and serological classification of inflammatory bowel disease: Report of a Working Party of the 2005 Montreal World Congress of Gastroenterology. Can J Gastroenterol 2005;(19 Suppl A):5e36. Satsangi J, Silverberg MS, Vermeire S, et al. The Montreal classification of inflammatory bowel disease: controversies, consensus, and implications. Gut 2006;55:749e53. Card T, Hubbard R, Logan RF. Mortality in inflammatory bowel disease: a population-based cohort study. Gastroenterology 2003;125:1583e90.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 26. 27.

28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58.

Langholz E, Munkholm P, Davidsen M, et al. Colorectal cancer risk and mortality in patients with ulcerative colitis. Gastroenterology 1992;103:1444e51. Jess T, Riis L, Vind I, et al. Changes in clinical characteristics, course, and prognosis of inflammatory bowel disease during the last 5 decades: a population-based study from Copenhagen, Denmark. Inflamm Bowel Dis 2007;13:481e9. Ekbom A, Helmick C, Zack M, et al. Ulcerative colitis and colorectal cancer. A population-based study. N Engl J Med 1990;323:1228e33. Ekbom A, Helmick C, Zack M, et al. Increased risk of large-bowel cancer in Crohn’s disease with colonic involvement. Lancet 1990;336:357e9. Friedman S, Rubin PH, Bodian C, et al. Screening and surveillance colonoscopy in chronic Crohn’s colitis: results of a surveillance program spanning 25 years. Clin Gastroenterol Hepatol 2008;6:993e8; quiz 53e4. Softley A, Clamp SE, Watkinson G, et al. The natural history of inflammatory bowel disease: has there been a change in the last 20 years? Scand J Gastroenterol Suppl 1988;144:20e3. Mahid SS, Minor KS, Soto RE, et al. Smoking and inflammatory bowel disease: a meta-analysis. Mayo Clin Proc 2006;81:1462e71. Sands BE. From symptom to diagnosis: clinical distinctions among various forms of intestinal inflammation. Gastroenterology 2004;126:1518e32. Nguyen GC, Kaplan GG, Harris ML, et al. A national survey of the prevalence and impact of Clostridium difficile infection among hospitalized inflammatory bowel disease patients. Am J Gastroenterol 2008;103:1443e50. Issa M, Ananthakrishnan AN, Binion DG. Clostridium difficile and inflammatory bowel disease. Inflamm Bowel Dis 2008;14:1432e42. Terheggen G, Lanyi B, Schanz S, et al. Safety, feasibility, and tolerability of ileocolonoscopy in inflammatory bowel disease. Endoscopy 2008;40:656e63. Bourreille A, Ignjatovic A, Aabakken L, et al. Role of small-bowel endoscopy in the management of patients with inflammatory bowel disease: an international OMED-ECCO consensus. Endoscopy 2009;41:618e37. Jenkins D, Balsitis M, Gallivan S, et al. Guidelines for the initial biopsy diagnosis of suspected chronic idiopathic inflammatory bowel disease. The British Society of Gastroenterology Initiative. J Clin Pathol 1997;50:93e105. Hyde GM, Jewell DP, Warren BF. Histological changes associated with the use of intravenous cyclosporin in the treatment of severe ulcerative colitis may mimic dysplasia. Colorectal Dis 2002;4:455e8. Brenner DJ, Hall EJ. Computed tomographydan increasing source of radiation exposure. N Engl J Med 2007;357:2277e84. Hall EJ, Brenner DJ. Cancer risks from diagnostic radiology. Br J Radiol 2008;81:362e78. Parente F, Maconi G, Bollani S, et al. Bowel ultrasound in assessment of Crohn’s disease and detection of related small bowel strictures: a prospective comparative study versus x ray and intraoperative findings. Gut 2002;50:490e5. Dietrich CF. Significance of abdominal ultrasound in inflammatory bowel disease. Dig Dis 2009;27:482e93. Gourtsoyiannis NC, Grammatikakis J, Papamastorakis G, et al. Imaging of small intestinal Crohn’s disease: comparison between MR enteroclysis and conventional enteroclysis. Eur Radiol 2006;16:1915e25. Albert JG, Martiny F, Krummenerl A, et al. Diagnosis of small bowel Crohn’s disease: a prospective comparison of capsule endoscopy with magnetic resonance imaging and fluoroscopic enteroclysis. Gut 2005;54:1721e7. Desmond AN, O’Regan K, Curran C, et al. Crohn’s disease: factors associated with exposure to high levels of diagnostic radiation. Gut 2008;57:1524e9. Stathaki MI, Koukouraki SI, Karkavitsas NS, et al. Role of scintigraphy in inflammatory bowel disease. World J Gastroenterol 2009;15:2693e700. Sailer J, Peloschek P, Schober E, et al. Diagnostic value of CT enteroclysis compared with conventional enteroclysis in patients with Crohn’s disease. AJR Am J Roentgenol 2005;185:1575e81. Martin DR, Lauenstein T, Sitaraman SV. Utility of magnetic resonance imaging in small bowel Crohn’s disease. Gastroenterology 2007;133:385e90. Stratton RJ, Hackston A, Longmore D, et al. Malnutrition in hospital outpatients and inpatients: prevalence, concurrent validity and ease of use of the ‘malnutrition universal screening tool’ (‘MUST’) for adults. Br J Nutr 2004;92:799e808. Headstrom PD, Rulyak SJ, Lee SD. Prevalence of and risk factors for vitamin B(12) deficiency in patients with Crohn’s disease. Inflamm Bowel Dis 2008;14:217e23. National Collaborating Centre for Acute Care. Nutrition Support in Adults Oral Nutrition Support, Enteral Tube Feeding and Parenteral Nutrition. London: National Collaborating Centre for Acute Care, 2006. http://www.rcseng.ac.uk. Bakker H, Bozzetti F, Staun M, et al. Home parenteral nutrition in adults: a european multicentre survey in 1997. ESPEN-Home Artificial Nutrition Working Group. Clin Nutr 1999;18:135e40. Nightingale J, Woodward JM. Guidelines for management of patients with a short bowel. Gut 2006;(55 Suppl 4):iv1e12. Lochs H, Dejong C, Hammarqvist F, et al. ESPEN guidelines on enteral nutrition: gastroenterology. Clin Nutr 2006;25:260e74. Fell JM, Paintin M, Arnaud-Battandier F, et al. Mucosal healing and a fall in mucosal pro-inflammatory cytokine mRNA induced by a specific oral polymeric diet in paediatric Crohn’s disease. Aliment Pharmacol Ther 2000;14:281e9. Sanderson IR, Udeen S, Davies PS, et al. Remission induced by an elemental diet in small bowel Crohn’s disease. Arch Dis Child 1987;62:123e7. Zachos M, Tondeur M, Griffiths AM. Enteral nutritional therapy for induction of remission in Crohn’s disease. Cochrane Database Syst Rev 2007;(1):CD000542.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86.

87.

88. 89. 90.

Akobeng AK, Thomas AG. Enteral nutrition for maintenance of remission in Crohn’s disease. Cochrane Database Syst Rev 2007;(3):CD005984. Heuschkel RB, Menache CC, Megerian JT, et al. Enteral nutrition and corticosteroids in the treatment of acute Crohn’s disease in children. J Pediatr Gastroenterol Nutr 2000;31:8e15. Looijer-van Langen MA, Dieleman LA. Prebiotics in chronic intestinal inflammation. Inflamm Bowel Dis 2009;15:454e62. Pham M, Lemberg DA, Day AS. Probiotics: sorting the evidence from the myths. Med J Aust 2008;188:304e8. Mimura T, Rizzello F, Helwig U, et al. Once daily high dose probiotic therapy (VSL#3) for maintaining remission in recurrent or refractory pouchitis. Gut 2004;53:108e14. Gionchetti P, Rizzello F, Helwig U, et al. Prophylaxis of pouchitis onset with probiotic therapy: a double-blind, placebo-controlled trial. Gastroenterology 2003;124:1202e9. Bibiloni R, Fedorak RN, Tannock GW, et al. VSL#3 probiotic-mixture induces remission in patients with active ulcerative colitis. Am J Gastroenterol 2005;100:1539e46. Miele E, Pascarella F, Giannetti E, et al. Effect of a probiotic preparation (VSL#3) on induction and maintenance of remission in children with ulcerative colitis. Am J Gastroenterol 2009;104:437e43. Schultz M. Clinical use of E. coli Nissle 1917 in inflammatory bowel disease. Inflamm Bowel Dis 2008;14:1012e18. Rolfe VE, Fortun PJ, Hawkey CJ, et al. Probiotics for maintenance of remission in Crohn’s disease. Cochrane Database Syst Rev 2006;(4):CD004826. Van Gossum A, Cabre E, Hebuterne X, et al. ESPEN guidelines on parenteral nutrition: gastroenterology. Clin Nutr 2009;28:415e27. Greenberg GR, Fleming CR, Jeejeebhoy KN, et al. Controlled trial of bowel rest and nutritional support in the management of Crohn’s disease. Gut 1988;29:1309e15. McIntyre PB, Powell-Tuck J, Wood SR, et al. Controlled trial of bowel rest in the treatment of severe acute colitis. Gut 1986;27:481e5. Payne-James JJ, Silk DB. Total parenteral nutrition as primary treatment in Crohn’s diseasedRIP? Gut 1988;29:1304e8. Calkins BM. A meta-analysis of the role of smoking in inflammatory bowel disease. Dig Dis Sci 1989;34:1841e54. Cosnes J, Carbonnel F, Beaugerie L, et al. Effects of cigarette smoking on the long-term course of Crohn’s disease. Gastroenterology 1996;110:424e31. Cottone M, Rosselli M, Orlando A, et al. Smoking habits and recurrence in Crohn’s disease. Gastroenterology 1994;106:643e8. Duffy LC, Zielezny MA, Marshall JR, et al. Cigarette smoking and risk of clinical relapse in patients with Crohn’s disease. Am J Prev Med 1990;6:161e6. Reese GE, Nanidis T, Borysiewicz C, et al. The effect of smoking after surgery for Crohn’s disease: a meta-analysis of observational studies. Int J Colorectal Dis 2008;23:1213e21. Sutherland LR, Ramcharan S, Bryant H, et al. Effect of cigarette smoking on recurrence of Crohn’s disease. Gastroenterology 1990;98:1123e8. Johnson GJ, Cosnes J, Mansfield JC. Review article: smoking cessation as primary therapy to modify the course of Crohn’s disease. Aliment Pharmacol Ther 2005;21:921e31. Kefalakes H, Stylianides TJ, Amanakis G, et al. Exacerbation of inflammatory bowel diseases associated with the use of nonsteroidal anti-inflammatory drugs: myth or reality? Eur J Clin Pharmacol 2009;65:963e70. Takeuchi K, Smale S, Premchand P, et al. Prevalence and mechanism of nonsteroidal anti-inflammatory drug-induced clinical relapse in patients with inflammatory bowel disease. Clin Gastroenterol Hepatol 2006;4:196e202. Sandborn WJ, Hanauer SB. Systematic review: the pharmacokinetic profiles of oral mesalazine formulations and mesalazine pro-drugs used in the management of ulcerative colitis. Aliment Pharmacol Ther 2003;17:29e42. Sutherland L, Macdonald JK. Oral 5-aminosalicylic acid for induction of remission in ulcerative colitis. Cochrane Database Syst Rev 2006;(2):CD000543. Bergman R, Parkes M. Systematic review: the use of mesalazine in inflammatory bowel disease. Aliment Pharmacol Ther 2006;23:841e55. Kamm MA, Sandborn WJ, Gassull M, et al. Once-daily, high-concentration MMX mesalamine in active ulcerative colitis. Gastroenterology 2007;132:66e75; quiz 432e3. Hanauer SB, Sandborn WJ, Kornbluth A, et al. Delayed-release oral mesalamine at 4.8 g/day (800 mg tablet) for the treatment of moderately active ulcerative colitis: the ASCEND II trial. Am J Gastroenterol 2005;100:2478e85. Marteau P, Probert CS, Lindgren S, et al. Combined oral and enema treatment with Pentasa (mesalazine) is superior to oral therapy alone in patients with extensive mild/moderate active ulcerative colitis: a randomised, double blind, placebo controlled study. Gut 2005;54:960e5. Kruis W, Kiudelis G, Racz I, et al. Once daily versus three times daily mesalazine granules in active ulcerative colitis: a double-blind, double-dummy, randomised, non-inferiority trial. Gut 2009;58:233e40. Marshall JK, Thabane M, Steinhart AH, et al. Rectal 5-aminosalicylic acid for induction of remission in ulcerative colitis. Cochrane Database Syst Rev 2010;(1): CD004115. Sutherland L, Macdonald JK. Oral 5-aminosalicylic acid for maintenance of remission in ulcerative colitis. Cochrane Database Syst Rev 2006;(2):CD000544.

601

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 91. 92. 93. 94. 95. 96. 97. 98.

99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116.

117. 118. 119. 120.

602

Dignass AU, Bokemeyer B, Adamek H, et al. Mesalamine once daily is more effective than twice daily in patients with quiescent ulcerative colitis. Clin Gastroenterol Hepatol 2009;7:762e9. Kamm MA, Lichtenstein GR, Sandborn WJ, et al. Randomised trial of once- or twice-daily MMX mesalazine for maintenance of remission in ulcerative colitis. Gut 2008;57:893e902. Eaden J, Abrams K, Ekbom A, et al. Colorectal cancer prevention in ulcerative colitis: a case-control study. Aliment Pharmacol Ther 2000;14:145e53. Hanauer SB, Stromberg U. Oral Pentasa in the treatment of active Crohn’s disease: a meta-analysis of double-blind, placebo-controlled trials. Clin Gastroenterol Hepatol 2004;2:379e88. Summers RW, Switz DM, Sessions JT Jr, et al. National Cooperative Crohn’s Disease Study: results of drug treatment. Gastroenterology 1979;77:847e69. Rasmussen SN, Lauritsen K, Tage-Jensen U, et al. 5-Aminosalicylic acid in the treatment of Crohn’s disease. A 16-week double-blind, placebo-controlled, multicentre study with Pentasa. Scand J Gastroenterol 1987;22:877e83. Mahida YR, Jewell DP. Slow-release 5-amino-salicylic acid (Pentasa) for the treatment of active Crohn’s disease. Digestion 1990;45:88e92. Tremaine WJ, Schroeder KW, Harrison JM, et al. A randomized, double-blind, placebo-controlled trial of the oral mesalamine (5-ASA) preparation, Asacol, in the treatment of symptomatic Crohn’s colitis and ileocolitis. J Clin Gastroenterol 1994;19:278e82. Singleton JW, Hanauer SB, Gitnick GL, et al. Mesalamine capsules for the treatment of active Crohn’s disease: results of a 16-week trial. Pentasa Crohn’s Disease Study Group. Gastroenterology 1993;104:1293e301. Singleton J. Second trial of mesalamine therapy in the treatment of active Crohn’s disease. Gastroenterology 1994;107:632e3. Singleton JW, Hanauer S, Robinson M. Quality-of-life results of double-blind, placebo-controlled trial of mesalamine in patients with Crohn’s disease. Dig Dis Sci 1995;40:931e5. Akobeng AK, Gardener E. Oral 5-aminosalicylic acid for maintenance of medicallyinduced remission in Crohn’s Disease. Cochrane Database Syst Rev 2005;(1): CD003715. Loftus EV Jr, Kane SV, Bjorkman D. Systematic review: short-term adverse effects of 5-aminosalicylic acid agents in the treatment of ulcerative colitis. Aliment Pharmacol Ther 2004;19:179e89. Ransford RA, Langman MJ. Sulphasalazine and mesalazine: serious adverse reactions re-evaluated on the basis of suspected adverse reaction reports to the Committee on Safety of Medicines. Gut 2002;51:536e9. Van Staa TP, Travis S, Leufkens HG, et al. 5-aminosalicylic acids and the risk of renal disease: a large British epidemiologic study. Gastroenterology 2004;126:1733e9. Muller AF, Stevens PE, McIntyre AS, et al. Experience of 5-aminosalicylate nephrotoxicity in the United Kingdom. Aliment Pharmacol Ther 2005;21:1217e24. Castiglione F, Rispo A, Di Girolamo E, et al. Antibiotic treatment of small bowel bacterial overgrowth in patients with Crohn’s disease. Aliment Pharmacol Ther 2003;18:1107e12. Blichfeldt P, Blomhoff JP, Myhre E, et al. Metronidazole in Crohn’s disease. A double blind cross-over clinical trial. Scand J Gastroenterol 1978;13:123e7. Sutherland L, Singleton J, Sessions J, et al. Double blind, placebo controlled trial of metronidazole in Crohn’s disease. Gut 1991;32:1071e5. Ursing B, Alm T, Barany F, et al. A comparative study of metronidazole and sulfasalazine for active Crohn’s disease: the cooperative Crohn’s disease study in Sweden. II. Result. Gastroenterology 1982;83:550e62. Rutgeerts P, Hiele M, Geboes K, et al. Controlled trial of metronidazole treatment for prevention of Crohn’s recurrence after ileal resection. Gastroenterology 1995;108:1617e21. Thia KT, Mahadevan U, Feagan BG, et al. Ciprofloxacin or metronidazole for the treatment of perianal fistulas in patients with Crohn’s disease: a randomized, double-blind, placebo-controlled pilot study. Inflamm Bowel Dis 2009;15:17e24. Madden MV, McIntyre AS, Nicholls RJ. Double-blind crossover trial of metronidazole versus placebo in chronic unremitting pouchitis. Dig Dis Sci 1994;39:1193e6. Brandt LJ, Bernstein LH, Boley SJ, et al. Metronidazole therapy for perineal Crohn’s disease: a follow-up study. Gastroenterology 1982;83:383e7. Prantera C, Zannoni F, Scribano ML, et al. An antibiotic regimen for the treatment of active Crohn’s disease: a randomized, controlled clinical trial of metronidazole plus ciprofloxacin. Am J Gastroenterol 1996;91:328e32. Colombel JF, Lemann M, Cassagnou M, et al. A controlled trial comparing ciprofloxacin with mesalazine for the treatment of active Crohn’s disease. Groupe d’Etudes Therapeutiques des Affections Inflammatoires Digestives (GETAID). Am J Gastroenterol 1999;94:674e8. Mantzaris GJ, Archavlis E, Christoforidis P, et al. A prospective randomized controlled trial of oral ciprofloxacin in acute ulcerative colitis. Am J Gastroenterol 1997;92:454e6. Turunen UM, Farkkila MA, Hakala K, et al. Long-term treatment of ulcerative colitis with ciprofloxacin: a prospective, double-blind, placebo-controlled study. Gastroenterology 1998;115:1072e8. Shen B, Achkar JP, Lashner BA, et al. A randomized clinical trial of ciprofloxacin and metronidazole to treat acute pouchitis. Inflamm Bowel Dis 2001;7:301e5. Seow CH, Benchimol EI, Steinhart AH, et al. Budesonide for Crohn’s disease. Expert Opin Drug Metab Toxicol 2009;5:971e9.

121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131.

132. 133. 134. 135.

136. 137. 138. 139.

140. 141. 142. 143. 144. 145. 146. 147. 148.

Campieri M, Adamo S, Valpiani D, et al. Oral beclometasone dipropionate in the treatment of extensive and left-sided active ulcerative colitis: a multicentre randomised study. Aliment Pharmacol Ther 2003;17:1471e80. Manguso F, Balzano A. Meta-analysis: the efficacy of rectal beclomethasone dipropionate vs. 5-aminosalicylic acid in mild to moderate distal ulcerative colitis. Aliment Pharmacol Ther 2007;26:21e9. Benchimol EI, Seow CH, Steinhart AH, et al. Traditional corticosteroids for induction of remission in Crohn’s disease. Cochrane Database Syst Rev 2008;(2): CD006792. Truelove SC, Watkinson G, Draper G. Comparison of corticosteroid and sulphasalazine therapy in ulcerative colitis. Br Med J 1962;2:1708e11. Baron JH, Connell AM, Kanaghinis TG, et al. Out-patient treatment of ulcerative colitis. Comparison between three doses of oral prednisone. Br Med J 1962;2:441e3. Lennard-Jones JE, Longmore AJ, Newell AC, et al. An assessment of prednisone, salazopyrin, and topical hydrocortisone hemisuccinate used as out-patient treatment for ulcerative colitis. Gut 1960;1:217e22. Lofberg R, Danielsson A, Suhr O, et al. Oral budesonide versus prednisolone in patients with active extensive and left-sided ulcerative colitis. Gastroenterology 1996;110:1713e18. Lee FI, Jewell DP, Mani V, et al. A randomised trial comparing mesalazine and prednisolone foam enemas in patients with acute distal ulcerative colitis. Gut 1996;38:229e33. Marshall JK, Irvine EJ. Rectal corticosteroids versus alternative treatments in ulcerative colitis: a meta-analysis. Gut 1997;40:775e81. Malchow H, Ewe K, Brandes JW, et al. European Cooperative Crohn’s Disease Study (ECCDS): results of drug treatment. Gastroenterology 1984;86:249e66. Modigliani R, Mary JY, Simon JF, et al. Clinical, biological, and endoscopic picture of attacks of Crohn’s disease. Evolution on prednisolone. Groupe d’Etude Therapeutique des Affections Inflammatoires Digestives. Gastroenterology 1990;98:811e18. Kane SV, Schoenfeld P, Sandborn WJ, et al. The effectiveness of budesonide therapy for Crohn’s disease. Aliment Pharmacol Ther 2002;16:1509e17. Munkholm P, Langholz E, Davidsen M, et al. Frequency of glucocorticoid resistance and dependency in Crohn’s disease. Gut 1994;35:360e2. Olaison G, Sjodahl R, Tagesson C. Glucocorticoid treatment in ileal Crohn’s disease: relief of symptoms but not of endoscopically viewed inflammation. Gut 1990;31:325e8. Mantzaris GJ, Christidou A, Sfakianakis M, et al. Azathioprine is superior to budesonide in achieving and maintaining mucosal healing and histologic remission in steroid-dependent Crohn’s disease. Inflamm Bowel Dis 2009;15:375e82. Subramanian V, Saxena S, Kang JY, et al. Preoperative steroid use and risk of postoperative complications in patients with inflammatory bowel disease undergoing abdominal surgery. Am J Gastroenterol 2008;103:2373e81. Lichtenstein GR, Feagan BG, Cohen RD, et al. Serious infections and mortality in association with therapies for Crohn’s disease: TREAT registry. Clin Gastroenterol Hepatol 2006;4:621e30. Newby EA, Sawczenko A, Thomas AG, et al. Interventions for growth failure in childhood Crohn’s disease. Cochrane Database Syst Rev 2005;(3):CD003873. Lichtenstein GR, Abreu MT, Cohen R, et al. American Gastroenterological Association Institute medical position statement on corticosteroids, immunomodulators, and infliximab in inflammatory bowel disease. Gastroenterology 2006;130:935e9. Tiede I, Fritz G, Strand S, et al. CD28-dependent Rac1 activation is the molecular target of azathioprine in primary human CD4+ T lymphocytes. J Clin Invest 2003;111:1133e45. McGovern DP, Travis SP, Duley J, et al. Azathioprine intolerance in patients with IBD may be imidazole-related and is independent of TPMT activity. Gastroenterology 2002;122:838e9. Bowen DG, Selby WS. Use of 6-mercaptopurine in patients with inflammatory bowel disease previously intolerant of azathioprine. Dig Dis Sci 2000;45: 1810e13. Ardizzone S, Maconi G, Russo A, et al. Randomised controlled trial of azathioprine and 5-aminosalicylic acid for treatment of steroid dependent ulcerative colitis. Gut 2006;55:47e53. Timmer A, McDonald JW, Macdonald JK. Azathioprine and 6-mercaptopurine for maintenance of remission in ulcerative colitis. Cochrane Database Syst Rev 2007; (1):CD000478. Prefontaine E, Macdonald JK, Sutherland LR. Azathioprine or 6-mercaptopurine for induction of remission in Crohn’s disease. Cochrane Database Syst Rev 2009;(4): CD000545. Prefontaine E, Sutherland LR, Macdonald JK, et al. Azathioprine or 6mercaptopurine for maintenance of remission in Crohn’s disease. Cochrane Database Syst Rev 2009;(1):CD000067. Gilissen LP, Bierau J, Derijks LJ, et al. The pharmacokinetic effect of discontinuation of mesalazine on mercaptopurine metabolite levels in inflammatory bowel disease patients. Aliment Pharmacol Ther 2005;22:605e11. de Boer NK, Wong DR, Jharap B, et al. Dose-dependent influence of 5-aminosalicylates on thiopurine metabolism. Am J Gastroenterol 2007;102:2747e53.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 149.

150. 151.

152. 153. 154. 155.

156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177.

Hande S, Wilson-Rich N, Bousvaros A, et al. 5-aminosalicylate therapy is associated with higher 6-thioguanine levels in adults and children with inflammatory bowel disease in remission on 6-mercaptopurine or azathioprine. Inflamm Bowel Dis 2006;12:251e7. Lennard L, Gibson BE, Nicole T, et al. Congenital thiopurine methyltransferase deficiency and 6-mercaptopurine toxicity during treatment for acute lymphoblastic leukaemia. Arch Dis Child 1993;69:577e9. Gearry RB, Barclay ML, Burt MJ, et al. Thiopurine S-methyltransferase (TPMT) genotype does not predict adverse drug reactions to thiopurine drugs in patients with inflammatory bowel disease. Aliment Pharmacol Ther 2003;18:395e400. Gisbert JP, Nino P, Rodrigo L, et al. Thiopurine methyltransferase (TPMT) activity and adverse effects of azathioprine in inflammatory bowel disease: long-term follow-up study of 394 patients. Am J Gastroenterol 2006;101:2769e76. Lennard L. TPMT in the treatment of Crohn’s disease with azathioprine. Gut 2002;51:143e6. Colombel JF, Ferrari N, Debuysere H, et al. Genotypic analysis of thiopurine S-methyltransferase in patients with Crohn’s disease and severe myelosuppression during azathioprine therapy. Gastroenterology 2000;118:1025e30. Thomas CW Jr, Lowry PW, Franklin CL, et al. Erythrocyte mean corpuscular volume as a surrogate marker for 6-thioguanine nucleotide concentration monitoring in patients with inflammatory bowel disease treated with azathioprine or 6-mercaptopurine. Inflamm Bowel Dis 2003;9:237e45. Gisbert JP, Gomollon F. Thiopurine-induced myelotoxicity in patients with inflammatory bowel disease: a review. Am J Gastroenterol 2008;103:1783e800. Connell WR, Kamm MA, Ritchie JK, et al. Bone marrow toxicity caused by azathioprine in inflammatory bowel disease: 27 years of experience. Gut 1993;34:1081e5. Kinlen LJ, Sheil AG, Peto J, et al. Collaborative United Kingdom-Australasian study of cancer in patients treated with immunosuppressive drugs. Br Med J 1979;2:1461e6. Askling J, Brandt L, Lapidus A, et al. Risk of haematopoietic cancer in patients with inflammatory bowel disease. Gut 2005;54:617e22. Lewis JD, Bilker WB, Brensinger C, et al. Inflammatory bowel disease is not associated with an increased risk of lymphoma. Gastroenterology 2001;121:1080e7. Masunaga Y, Ohno K, Ogawa R, et al. Meta-analysis of risk of malignancy with immunosuppressive drugs in inflammatory bowel disease. Ann Pharmacother 2007;41:21e8. Kandiel A, Fraser AG, Korelitz BI, et al. Increased risk of lymphoma among inflammatory bowel disease patients treated with azathioprine and 6-mercaptopurine. Gut 2005;54:1121e5. Beaugerie L, Brousse N, Bouvier AM, et al. Lymphoproliferative disorders in patients receiving thiopurines for inflammatory bowel disease: a prospective observational cohort study. Lancet 2009;374:1617e25. Lewis JD, Schwartz JS, Lichtenstein GR. Azathioprine for maintenance of remission in Crohn’s disease: benefits outweigh the risk of lymphoma. Gastroenterology 2000;118:1018e24. Austin AS, Spiller RC. Inflammatory bowel disease, azathioprine and skin cancer: case report and literature review. Eur J Gastroenterol Hepatol 2001;13:193e4. Subramanian V, Pollok RC, Kang JY, et al. Systematic review of postoperative complications in patients with inflammatory bowel disease treated with immunomodulators. Br J Surg 2006;93:793e9. Myrelid P, Olaison G, Sjodahl R, et al. Thiopurine therapy is associated with postoperative intra-abdominal septic complications in abdominal surgery for Crohn’s disease. Dis Colon Rectum 2009;52:1387e94. Fraser AG, Orchard TR, Jewell DP. The efficacy of azathioprine for the treatment of inflammatory bowel disease: a 30 year review. Gut 2002;50:485e9. Lemann M, Mary JY, Colombel JF, et al. A randomized, double-blind, controlled withdrawal trial in Crohn’s disease patients in long-term remission on azathioprine. Gastroenterology 2005;128:1812e18. Treton X, Bouhnik Y, Mary JY, et al. Azathioprine withdrawal in patients with Crohn’s disease maintained on prolonged remission: a high risk of relapse. Clin Gastroenterol Hepatol 2009;7:80e5. Kim PS, Zlatanic J, Korelitz BI, et al. Optimum duration of treatment with 6-mercaptopurine for Crohn’s disease. Am J Gastroenterol 1999;94:3254e7. Cassinotti A, Actis GC, Duca P, et al. Maintenance Treatment With Azathioprine in Ulcerative Colitis: Outcome and Predictive Factors After Drug Withdrawal. Am J Gastroenterol 2009;104:2760e7. Alfadhli AA, McDonald JW, Feagan BG. Methotrexate for induction of remission in refractory Crohn’s disease. Cochrane Database Syst Rev 2005;(1):CD003459. Patel V, Macdonald JK, McDonald JW, et al. Methotrexate for maintenance of remission in Crohn’s disease. Cochrane Database Syst Rev 2009;(4):CD006884. Kozarek RA, Patterson DJ, Gelfand MD, et al. Methotrexate induces clinical and histologic remission in patients with refractory inflammatory bowel disease. Ann Intern Med 1989;110:353e6. Manosa M, Naves JE, Leal C, et al. Does methotrexate induce mucosal healing in Crohn’s disease? Inflamm Bowel Dis 2010;16:377e8. Feagan BG, Rochon J, Fedorak RN, et al. Methotrexate for the treatment of Crohn’s disease. The North American Crohn’s Study Group Investigators. N Engl J Med 1995;332:292e7.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

178. 179. 180. 181. 182.

183. 184. 185. 186.

187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199.

200. 201. 202. 203. 204. 205. 206. 207.

Feagan BG, Fedorak RN, Irvine EJ, et al. A comparison of methotrexate with placebo for the maintenance of remission in Crohn’s disease. North American Crohn’s Study Group Investigators. N Engl J Med 2000;342:1627e32. Oren R, Arber N, Odes S, et al. Methotrexate in chronic active ulcerative colitis: a double-blind, randomized, Israeli multicenter trial. Gastroenterology 1996;110:1416e21. Ei-Matary W, Vandermeer B, Griffiths AM. Methotrexate for maintenance of remission in ulcerative colitis. Cochrane Database Syst Rev 2009;(3):CD007560. Fraser AG, Morton D, McGovern D, et al. The efficacy of methotrexate for maintaining remission in inflammatory bowel disease. Aliment Pharmacol Ther 2002;16:693e7. Paoluzi OA, Pica R, Marcheggiano A, et al. Azathioprine or methotrexate in the treatment of patients with steroid-dependent or steroid-resistant ulcerative colitis: results of an open-label study on efficacy and tolerability in inducing and maintaining remission. Aliment Pharmacol Ther 2002;16:1751e9. Cummings JR, Herrlinger KR, Travis SP, et al. Oral methotrexate in ulcerative colitis. Aliment Pharmacol Ther 2005;21:385e9. Wahed M, Louis-Auguste JR, Baxter LM, et al. Efficacy of methotrexate in Crohn’s disease and ulcerative colitis patients unresponsive or intolerant to azathioprine /mercaptopurine. Aliment Pharmacol Ther 2009;30:614e20. Jundt JW, Browne BA, Fiocco GP, et al. A comparison of low dose methotrexate bioavailability: oral solution, oral tablet, subcutaneous and intramuscular dosing. J Rheumatol 1993;20:1845e9. Braun J, Kastner P, Flaxenberg P, et al. Comparison of the clinical efficacy and safety of subcutaneous versus oral administration of methotrexate in patients with active rheumatoid arthritis: results of a six-month, multicenter, randomized, doubleblind, controlled, phase IV trial. Arthritis Rheum 2008;58:73e81. Din S, Dahele A, Fennel J, et al. Use of methotrexate in refractory Crohn’s disease: the Edinburgh experience. Inflamm Bowel Dis 2008;14:756e62. Chong RY, Hanauer SB, Cohen RD. Efficacy of parenteral methotrexate in refractory Crohn’s disease. Aliment Pharmacol Ther 2001;15:35e44. Nathan DM, Iser JH, Gibson PR. A single center experience of methotrexate in the treatment of Crohn’s disease and ulcerative colitis: a case for subcutaneous administration. J Gastroenterol Hepatol 2008;23:954e8. Fraser AG. Methotrexate: first-line or second-line immunomodulator? Eur J Gastroenterol Hepatol 2003;15:225e31. Te HS, Schiano TD, Kuan SF, et al. Hepatic effects of long-term methotrexate use in the treatment of inflammatory bowel disease. Am J Gastroenterol 2000;95:3150e6. Mahadevan U, Kane S. American gastroenterological association institute technical review on the use of gastrointestinal medications in pregnancy. Gastroenterology 2006;131:283e311. Hausmann J, Zabel K, Herrmann E, et al. Methotrexate for maintenance of remission in chronic active Crohn’s disease: Long-term single-center experience and meta-analysis of observational studies. Inflamm Bowel Dis 2010;16:1195e202. D’Haens G, Lemmens L, Geboes K, et al. Intravenous cyclosporine versus intravenous corticosteroids as single therapy for severe attacks of ulcerative colitis. Gastroenterology 2001;120:1323e9. Lichtiger S, Present DH, Kornbluth A, et al. Cyclosporine in severe ulcerative colitis refractory to steroid therapy. N Engl J Med 1994;330:1841e5. Shibolet O, Regushevskaya E, Brezis M, et al. Cyclosporine A for induction of remission in severe ulcerative colitis. Cochrane Database Syst Rev 2005;(1): CD004277. McDonald JW, Feagan BG, Jewell D, et al. Cyclosporine for induction of remission in Crohn’s disease. Cochrane Database Syst Rev 2005;(2):CD000297. Van Assche G, D’Haens G, Noman M, et al. Randomized, double-blind comparison of 4 mg/kg versus 2 mg/kg intravenous cyclosporine in severe ulcerative colitis. Gastroenterology 2003;125:1025e31. Actis GC, Aimo G, Priolo G, et al. Efficacy and efficiency of oral microemulsion cyclosporin versus intravenous and soft gelatin capsule cyclosporin in the treatment of severe steroid-refractory ulcerative colitis: an open-label retrospective trial. Inflamm Bowel Dis 1998;4:276e9. Moskovitz DN, Van Assche G, Maenhout B, et al. Incidence of colectomy during long-term follow-up after cyclosporine-induced remission of severe ulcerative colitis. Clin Gastroenterol Hepatol 2006;4:760e5. Marion JF, Present D. 6-MP maintains cyclosporine-induced response in patients with severe acute colitis. Am J Gastroenterol 1996;91:1975. Fernandez-Banares F, Bertran X, Esteve-Comas M, et al. Azathioprine is useful in maintaining long-term remission induced by intravenous cyclosporine in steroidrefractory severe ulcerative colitis. Am J Gastroenterol 1996;91:2498e9. Actis GC, Bresso F, Astegiano M, et al. Safety and efficacy of azathioprine in the maintenance of ciclosporin-induced remission of ulcerative colitis. Aliment Pharmacol Ther 2001;15:1307e11. Cohen RD, Stein R, Hanauer SB. Intravenous cyclosporin in ulcerative colitis: a fiveyear experience. Am J Gastroenterol 1999;94:1587e92. Ogata H, Matsui T, Nakamura M, et al. A randomised dose finding study of oral tacrolimus (FK506) therapy in refractory ulcerative colitis. Gut 2006;55:1255e62. Herrlinger KR, Fellermann K, Stange EF. Tacrolimusefinally! Gut 2006;55:1224e5. Yamamoto S, Nakase H, Mikami S, et al. Long-term effect of tacrolimus therapy in patients with refractory ulcerative colitis. Aliment Pharmacol Ther 2008;28: 589e97.

603

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236.

237.

604

Fellermann K, Tanko Z, Herrlinger KR, et al. Response of refractory colitis to intravenous or oral tacrolimus (FK506). Inflamm Bowel Dis 2002;8:317e24. Bousvaros A, Kirschner BS, Werlin SL, et al. Oral tacrolimus treatment of severe colitis in children. J Pediatr 2000;137:794e9. Hanauer SB, Feagan BG, Lichtenstein GR, et al. Maintenance infliximab for Crohn’s disease: the ACCENT I randomised trial. Lancet 2002;359:1541e9. Rutgeerts P, Van Assche G, Vermeire S. Optimizing anti-TNF treatment in inflammatory bowel disease. Gastroenterology 2004;126:1593e610. Targan SR, Hanauer SB, van Deventer SJ, et al. A short-term study of chimeric monoclonal antibody cA2 to tumor necrosis factor alpha for Crohn’s disease. Crohn’s Disease cA2 Study Group. N Engl J Med 1997;337:1029e35. Hanauer SB, Sandborn WJ, Rutgeerts P, et al. Human anti-tumor necrosis factor monoclonal antibody (adalimumab) in Crohn’s disease: the CLASSIC-I trial. Gastroenterology 2006;130:323e33. Sandborn WJ, Rutgeerts P, Enns R, et al. Adalimumab induction therapy for Crohn disease previously treated with infliximab: a randomized trial. Ann Intern Med 2007;146:829e38. Colombel JF, Sandborn WJ, Rutgeerts P, et al. Adalimumab for maintenance of clinical response and remission in patients with Crohn’s disease: the CHARM trial. Gastroenterology 2007;132:52e65. Sandborn WJ, Hanauer SB, Rutgeerts P, et al. Adalimumab for maintenance treatment of Crohn’s disease: results of the CLASSIC II trial. Gut 2007;56:1232e9. Present DH, Rutgeerts P, Targan S, et al. Infliximab for the treatment of fistulas in patients with Crohn’s disease. N Engl J Med 1999;340:1398e405. Sands BE, Anderson FH, Bernstein CN, et al. Infliximab maintenance therapy for fistulizing Crohn’s disease. N Engl J Med 2004;350:876e85. Lawson MM, Thomas AG, Akobeng AK. Tumour necrosis factor alpha blocking agents for induction of remission in ulcerative colitis. Cochrane Database Syst Rev 2006;(3):CD005112. Rutgeerts P, Sandborn WJ, Feagan BG, et al. Infliximab for induction and maintenance therapy for ulcerative colitis. N Engl J Med 2005;353:2462e76. Sandborn WJ, Rutgeerts P, Feagan BG, et al. Colectomy rate comparison after treatment of ulcerative colitis with placebo or infliximab. Gastroenterology 2009;137:1250e60; quiz 520. Afif W, Leighton JA, Hanauer SB, et al. Open-label study of adalimumab in patients with ulcerative colitis including those with prior loss of response or intolerance to infliximab. Inflamm Bowel Dis 2009;15:1302e7. Barreiro-de Acosta M, Lorenzo A, Dominguez-Munoz JE. Adalimumab in ulcerative colitis: two cases of mucosal healing and clinical response at two years. World J Gastroenterol 2009;15:3814e16. Oussalah A, Laclotte C, Chevaux JB, et al. Long-term outcome of adalimumab therapy for ulcerative colitis with intolerance or lost response to infliximab: a singlecentre experience. Aliment Pharmacol Ther 2008;28:966e72. Peyrin-Biroulet L, Laclotte C, Roblin X, et al. Adalimumab induction therapy for ulcerative colitis with intolerance or lost response to infliximab: an open-label study. World J Gastroenterol 2007;13:2328e32. Probert CS, Hearing SD, Schreiber S, et al. Infliximab in moderately severe glucocorticoid resistant ulcerative colitis: a randomised controlled trial. Gut 2003;52:998e1002. Jarnerot G, Hertervig E, Friis-Liby I, et al. Infliximab as rescue therapy in severe to moderately severe ulcerative colitis: a randomized, placebo-controlled study. Gastroenterology 2005;128:1805e11. Lindgren SC, Flood LM, Kilander AF, et al. Early predictors of glucocorticosteroid treatment failure in severe and moderately severe attacks of ulcerative colitis. Eur J Gastroenterol Hepatol 1998;10:831e5. Toruner M, Loftus EV Jr, Harmsen WS, et al. Risk factors for opportunistic infections in patients with inflammatory bowel disease. Gastroenterology 2008;134:929e36. Lees CW, Heys D, Ho GT, et al. A retrospective analysis of the efficacy and safety of infliximab as rescue therapy in acute severe ulcerative colitis. Aliment Pharmacol Ther 2007;26:411e19. Selvasekar CR, Cima RR, Larson DW, et al. Effect of infliximab on short-term complications in patients undergoing operation for chronic ulcerative colitis. J Am Coll Surg 2007;204:956e62; discussion 62e3. Mor IJ, Vogel JD, da Luz Moreira A, et al. Infliximab in ulcerative colitis is associated with an increased risk of postoperative complications after restorative proctocolectomy. Dis Colon Rectum 2008;51:1202e7; discussion 7e10. Ferrante M, D’Hoore A, Vermeire S, et al. Corticosteroids but not infliximab increase short-term postoperative infectious complications in patients with ulcerative colitis. Inflamm Bowel Dis 2009;15:1062e70. Yang Z, Wu Q, Wu K, Fan D. Meta-analysis: pre-operative infliximab treatment and short-term post-operative complications in patients with ulcerative colitis. Aliment Pharmacol Ther;31:486e92. Keane J, Gershon S, Wise RP, et al. Tuberculosis associated with infliximab, a tumor necrosis factor alpha-neutralizing agent. N Engl J Med 2001;345:1098e104. British Thoracic Society Standards of Care Committee. BTS recommendations for assessing risk and for managing Mycobacterium tuberculosis infection and disease in patients due to start anti-TNF-alpha treatment. Thorax 2005;60:800e5. Theis VS, Rhodes JM. Review article: minimizing tuberculosis during anti-tumour necrosis factor-alpha treatment of inflammatory bowel disease. Aliment Pharmacol Ther 2008;27:19e30.

238. 239. 240. 241. 242. 243. 244. 245. 246.

247. 248. 249. 250. 251. 252. 253. 254. 255.

256. 257. 258. 259. 260. 261.

262. 263. 264. 265. 266.

Lalvani A, Millington KA. Screening for tuberculosis infection prior to initiation of anti-TNF therapy. Autoimmun Rev 2008;8:147e52. Esteve M, Saro C, Gonzalez-Huix F, et al. Chronic hepatitis B reactivation following infliximab therapy in Crohn’s disease patients: need for primary prophylaxis. Gut 2004;53:1363e5. Baert F, Noman M, Vermeire S, et al. Influence of immunogenicity on the long-term efficacy of infliximab in Crohn’s disease. N Engl J Med 2003;348:601e8. Bender NK, Heilig CE, Droll B, et al. Immunogenicity, efficacy and adverse events of adalimumab in RA patients. Rheumatol Int 2007;27:269e74. West RL, Zelinkova Z, Wolbink GJ, et al. Immunogenicity negatively influences the outcome of adalimumab treatment in Crohn’s disease. Aliment Pharmacol Ther 2008;28:1122e6. Maser EA, Villela R, Silverberg MS, et al. Association of trough serum infliximab to clinical outcome after scheduled maintenance treatment for Crohn’s disease. Clin Gastroenterol Hepatol 2006;4:1248e54. Seow CH, Newman A, Irwin SP, et al. Trough serum infliximab: a predictive factor of clinical outcome for infliximab treatment in acute ulcerative colitis. Gut 2010;59:49e54. Rutgeerts P, Vermeire S, Van Assche G. Predicting the response to infliximab from trough serum levels. Gut 2010;59:7e8. Bongartz T, Sutton AJ, Sweeting MJ, et al. Anti-TNF antibody therapy in rheumatoid arthritis and the risk of serious infections and malignancies: systematic review and meta-analysis of rare harmful effects in randomized controlled trials. JAMA 2006;295:2275e85. Fidder H, Schnitzler F, Ferrante M, et al. Long-term safety of infliximab for the treatment of inflammatory bowel disease: a single-centre cohort study. Gut 2009;58:501e8. Russo EA, Harris AW, Campbell S, et al. Experience of maintenance infliximab therapy for refractory ulcerative colitis from six centres in England. Aliment Pharmacol Ther 2009;29:308e14. Colombel JF, Loftus EV Jr, Tremaine WJ, et al. The safety profile of infliximab in patients with Crohn’s disease: the Mayo clinic experience in 500 patients. Gastroenterology 2004;126:19e31. Lees CW, Ali AI, Thompson AI, et al. The safety profile of anti-tumour necrosis factor therapy in inflammatory bowel disease in clinical practice: analysis of 620 patient-years follow-up. Aliment Pharmacol Ther 2009;29:286e97. Lees CW, Ironside J, Wallace WA, et al. Resolution of non-small-cell lung cancer after withdrawal of anti-TNF therapy. N Engl J Med 2008;359:320e1. Rennard SI, Fogarty C, Kelsen S, et al. The safety and efficacy of infliximab in moderate to severe chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2007;175:926e34. Mackey AC, Green L, Liang LC, et al. Hepatosplenic T cell lymphoma associated with infliximab use in young patients treated for inflammatory bowel disease. J Pediatr Gastroenterol Nutr 2007;44:265e7. Shale M, Kanfer E, Panaccione R, et al. Hepatosplenic T cell lymphoma in inflammatory bowel disease. Gut 2008;57:1639e41. Siegel CA, Marden SM, Persing SM, et al. Risk of lymphoma associated with combination anti-tumor necrosis factor and immunomodulator therapy for the treatment of Crohn’s disease: a meta-analysis. Clin Gastroenterol Hepatol 2009;7:874e81. The Lenercept Multiple Sclerosis Study Group and The University of British Columbia MS/MRI Analysis Group. TNF neutralization in MS: results of a randomized, placebo-controlled multicenter study. Neurology 1999;53:457e65. D’Haens G, Baert F, van Assche G, et al. Early combined immunosuppression or conventional management in patients with newly diagnosed Crohn’s disease: an open randomised trial. Lancet 2008;371:660e7. Colombel JF, Sandborn WJ, Reinisch W, et al. Infliximab, azathioprine, or combination therapy for Crohn’s disease. N Engl J Med 2010;362:1383e95. Van Assche G, Magdelaine-Beuzelin C, D’Haens G, et al. Withdrawal of immunosuppression in Crohn’s disease treated with scheduled infliximab maintenance: a randomized trial. Gastroenterology 2008;134:1861e8. Oussalah A, Chevaux JB, Fay R, et al. Predictors of infliximab failure after azathioprine withdrawal in Crohn’s disease treated with combination therapy. Am J Gastroenterol 2010;105:1142e9. Ho GT, Mowat A, Potts L, et al. Efficacy and complications of adalimumab treatment for medically-refractory Crohn’s disease: analysis of nationwide experience in Scotland (2004e2008). Aliment Pharmacol Ther 2009;29:527e34. Carter JD, Ladhani A, Ricca LR, et al. A safety assessment of tumor necrosis factor antagonists during pregnancy: a review of the Food and Drug Administration database. J Rheumatol 2009;36:635e41. Truelove SC, Witts LJ. Cortisone in ulcerative colitis; final report on a therapeutic trial. Br Med J 1955;2:1041e8. Schroeder KW, Tremaine WJ, Ilstrup DM. Coated oral 5-aminosalicylic acid therapy for mildly to moderately active ulcerative colitis. A randomized study. N Engl J Med 1987;317:1625e9. Jewell DP, Truelove SC. Azathioprine in ulcerative colitis: final report on controlled therapeutic trial. Br Med J 1974;4:627e30. Cohen RD, Woseth DM, Thisted RA, et al. A meta-analysis and overview of the literature on treatment options for left-sided ulcerative colitis and ulcerative proctitis. Am J Gastroenterol 2000;95:1263e76.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 267. 268.

269. 270. 271. 272. 273. 274. 275. 276. 277. 278. 279. 280. 281. 282. 283. 284. 285. 286. 287. 288. 289. 290. 291. 292. 293. 294. 295. 296.

Safdi M, DeMicco M, Sninsky C, et al. A double-blind comparison of oral versus rectal mesalamine versus combination therapy in the treatment of distal ulcerative colitis. Am J Gastroenterol 1997;92:1867e71. Hanauer SB. Dose-ranging study of mesalamine (PENTASA) enemas in the treatment of acute ulcerative proctosigmoiditis: results of a multicentered placebocontrolled trial. The U.S. PENTASA Enema Study Group. Inflamm Bowel Dis 1998;4:79e83. Allison MC, Vallance R. Prevalence of proximal faecal stasis in active ulcerative colitis. Gut 1991;32:179e82. Travis SP, Farrant JM, Ricketts C, et al. Predicting outcome in severe ulcerative colitis. Gut 1996;38:905e10. Rosenberg W, Ireland A, Jewell DP. High-dose methylprednisolone in the treatment of active ulcerative colitis. J Clin Gastroenterol 1990;12:40e1. Turner D, Walsh CM, Steinhart AH, et al. Response to corticosteroids in severe ulcerative colitis: a systematic review of the literature and a meta-regression. Clin Gastroenterol Hepatol 2007;5:103e10. Gonzalez-Huix F, Fernandez-Banares F, Esteve-Comas M, et al. Enteral versus parenteral nutrition as adjunct therapy in acute ulcerative colitis. Am J Gastroenterol 1993;88:227e32. Kishore J, Ghoshal U, Ghoshal UC, et al. Infection with cytomegalovirus in patients with inflammatory bowel disease: prevalence, clinical significance and outcome. J Med Microbiol 2004;53:1155e60. Minami M, Ohta M, Ohkura T, et al. Cytomegalovirus infection in severe ulcerative colitis patients undergoing continuous intravenous cyclosporine treatment in Japan. World J Gastroenterol 2007;13:754e60. Papadakis KA, Tung JK, Binder SW, et al. Outcome of cytomegalovirus infections in patients with inflammatory bowel disease. Am J Gastroenterol 2001;96:2137e42. Rahier JF, Ben-Horin S, Chowers Y, et al. European evidenced based consensus on the prevention, diagnosis and management of opportunistic infections in inflammatory bowel disease. J Crohns Colitis 2009;3:47e91. Irving PM, Pasi KJ, Rampton DS. Thrombosis and inflammatory bowel disease. Clin Gastroenterol Hepatol 2005;3:617e28. Bojic D, Radojicic Z, Nedeljkovic-Protic M, et al. Long-term outcome after admission for acute severe ulcerative colitis in Oxford: the 1992e1993 cohort. Inflamm Bowel Dis 2009;15:823e8. Homik J, Suarez-Almazor ME, Shea B, et al. Calcium and vitamin D for corticosteroid-induced osteoporosis. Cochrane Database Syst Rev 2000;(2): CD000952. Hoie O, Wolters FL, Riis L, et al. Low colectomy rates in ulcerative colitis in an unselected European cohort followed for 10 years. Gastroenterology 2007;132:507e15. Alves A, Panis Y, Bouhnik Y, et al. Subtotal colectomy for severe acute colitis: a 20-year experience of a tertiary care center with an aggressive and early surgical policy. J Am Coll Surg 2003;197:379e85. Brown SR, Haboubi N, Hampton J, et al. The management of acute severe colitis: ACPGBI position statement. Colorectal Dis 2008;(10 Suppl 3):8e29. Seidel SA, Newman M, Sharp KW. Ileoanal pouch versus ileostomy: is there a difference in quality of life? Am Surg 2000;66:540e6; discussion 546e7. Singh B, Mortensen N, Shorthouse AJ. Defunctioning ileostomy following restorative proctocolectomy. Ann R Coll Surg Engl 2008;90: 541e5. Heuschen UA, Hinz U, Allemeyer EH, et al. One- or two-stage procedure for restorative proctocolectomy: rationale for a surgical strategy in ulcerative colitis. Ann Surg 2001;234:788e94. Remzi FH, Fazio VW, Gorgun E, et al. The outcome after restorative proctocolectomy with or without defunctioning ileostomy. Dis Colon Rectum 2006;49:470e7. Keighley MR, Yoshioka K, Kmiot W. Prospective randomized trial to compare the stapled double lumen pouch and the sutured quadruple pouch for restorative proctocolectomy. Br J Surg 1988;75:1008e11. Lovegrove RE, Constantinides VA, Heriot AG, et al. A comparison of hand-sewn versus stapled ileal pouch anal anastomosis (IPAA) following proctocolectomy: a meta-analysis of 4183 patients. Ann Surg 2006;244:18e26. Berndtsson I, Oresland T. Quality of life before and after proctocolectomy and IPAA in patients with ulcerative proctocolitisda prospective study. Colorectal Dis 2003;5:173e9. Fazio VW, Ziv Y, Church JM, et al. Ileal pouch-anal anastomoses complications and function in 1005 patients. Ann Surg 1995;222:120e7. Heuschen UA, Hinz U, Allemeyer EH, et al. Risk factors for ileoanal J pouch-related septic complications in ulcerative colitis and familial adenomatous polyposis. Ann Surg 2002;235:207e16. Delaini GG, Scaglia M, Colucci G, et al. The ileoanal pouch procedure in the longterm perspective: a critical review. Tech Coloproctol 2005;9:187e92. Fazio VW, Tekkis PP, Remzi F, et al. Quantification of risk for pouch failure after ileal pouch anal anastomosis surgery. Ann Surg 2003;238:605e14; discussion 614e17. Fazio VW, O’Riordain MG, Lavery IC, et al. Long-term functional outcome and quality of life after stapled restorative proctocolectomy. Ann Surg 1999;230:575e84; discussion 584e6. Tulchinsky H, Hawley PR, Nicholls J. Long-term failure after restorative proctocolectomy for ulcerative colitis. Ann Surg 2003;238:229e34.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

297. 298. 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. 310. 311. 312. 313. 314. 315. 316. 317. 318. 319. 320. 321. 322. 323. 324.

325. 326.

Tekkis PP, Senagore AJ, Delaney CP, et al. Evaluation of the learning curve in laparoscopic colorectal surgery: comparison of right-sided and left-sided resections. Ann Surg 2005;242:83e91. Kaplan GG, McCarthy EP, Ayanian JZ, et al. Impact of hospital volume on postoperative morbidity and mortality following a colectomy for ulcerative colitis. Gastroenterology 2008;134:680e7. Raval MJ, Schnitzler M, O’Connor BI, et al. Improved outcome due to increased experience and individualized management of leaks after ileal pouch-anal anastomosis. Ann Surg 2007;246:763e70. Hyman NH, Cataldo P, Osler T. Urgent subtotal colectomy for severe inflammatory bowel disease. Dis Colon Rectum 2005;48:70e3. Aberra FN, Lewis JD, Hass D, et al. Corticosteroids and immunomodulators: postoperative infectious complication risk in inflammatory bowel disease patients. Gastroenterology 2003;125:320e7. Poritz LS, Rowe WA, Swenson BR, et al. Intravenous cyclosporine for the treatment of severe steroid refractory ulcerative colitis: what is the cost? Dis Colon Rectum 2005;48:1685e90. Ferrante M, Declerck S, De Hertogh G, et al. Outcome after proctocolectomy with ileal pouch-anal anastomosis for ulcerative colitis. Inflamm Bowel Dis 2008;14:20e8. Michelassi F, Lee J, Rubin M, et al. Long-term functional results after ileal pouch anal restorative proctocolectomy for ulcerative colitis: a prospective observational study. Ann Surg 2003;238:433e41; discussion 442e5. Tekkis PP, Lovegrove RE, Tilney HS, et al. Long-term failure and function after restorative proctocolectomyda multi-centre study of patients from the uk national ileal pouch registry. Colorectal Dis 2010;12:433e41. Johnson P, Richard C, Ravid A, et al. Female infertility after ileal pouch-anal anastomosis for ulcerative colitis. Dis Colon Rectum 2004;47:1119e26. Gorgun E, Remzi FH, Goldberg JM, et al. Fertility is reduced after restorative proctocolectomy with ileal pouch anal anastomosis: a study of 300 patients. Surgery 2004;136:795e803. Ording Olsen K, Juul S, Berndtsson I, et al. Ulcerative colitis: female fecundity before diagnosis, during disease, and after surgery compared with a population sample. Gastroenterology 2002;122:15e19. Sagar PM, Pemberton JH. Ileo-anal pouch function and dysfunction. Dig Dis 1997;15:172e88. Shen B, Achkar JP, Lashner BA, et al. Endoscopic and histologic evaluation together with symptom assessment are required to diagnose pouchitis. Gastroenterology 2001;121:261e7. Bell AJ, Price AB, Forbes A, et al. Pre-pouch ileitis: a disease of the ileum in ulcerative colitis after restorative proctocolectomy. Colorectal Dis 2006;8:402e10. Shen B, Achkar JP, Lashner BA, et al. Irritable pouch syndrome: a new category of diagnosis for symptomatic patients with ileal pouch-anal anastomosis. Am J Gastroenterol 2002;97:972e7. Shen B, Lashner BA, Bennett AE, et al. Treatment of rectal cuff inflammation (cuffitis) in patients with ulcerative colitis following restorative proctocolectomy and ileal pouch-anal anastomosis. Am J Gastroenterol 2004;99:1527e31. Gionchetti P, Rizzello F, Morselli C, et al. High-dose probiotics for the treatment of active pouchitis. Dis Colon Rectum 2007;50:2075e82; discussion 2082e4. Gionchetti P, Rizzello F, Venturi A, et al. Oral bacteriotherapy as maintenance treatment in patients with chronic pouchitis: a double-blind, placebo-controlled trial. Gastroenterology 2000;119:305e9. Mimura T, Rizzello F, Helwig U, et al. Four-week open-label trial of metronidazole and ciprofloxacin for the treatment of recurrent or refractory pouchitis. Aliment Pharmacol Ther 2002;16:909e17. Sandborn WJ, McLeod R, Jewell DP. Medical therapy for induction and maintenance of remission in pouchitis: a systematic review. Inflamm Bowel Dis 1999;5:33e9. Gionchetti P, Rizzello F, Poggioli G, et al. Oral budesonide in the treatment of chronic refractory pouchitis. Aliment Pharmacol Ther 2007;25:1231e6. Sambuelli A, Boerr L, Negreira S, et al. Budesonide enema in pouchitisea double-blind, double-dummy, controlled trial. Aliment Pharmacol Ther 2002;16:27e34. Winter TA, Dalton HR, Merrett MN, et al. Cyclosporin A retention enemas in refractory distal ulcerative colitis and ’pouchitis’. Scand J Gastroenterol 1993;28:701e4. de Silva HJ, Ireland A, Kettlewell M, et al. Short-chain fatty acid irrigation in severe pouchitis. N Engl J Med 1989;321:1416e17. Ferrante M, D’Haens G, Dewit O, et al. Efficacy of infliximab in refractory pouchitis and Crohn’s disease-related complications of the pouch: A Belgian case series. Inflamm Bowel Dis 2010;16:243e9. Holubar SD, Cima RR, Sandborn WJ, et al. Treatment and prevention of pouchitis after ileal pouch-anal anastomosis for chronic ulcerative colitis. Cochrane Database Syst Rev 2010;(6):CD001176. Alves A, Panis Y, Bouhnik Y, et al. Risk factors for intra-abdominal septic complications after a first ileocecal resection for Crohn’s disease: a multivariate analysis in 161 consecutive patients. Dis Colon Rectum 2007;50:331e6. Hulten L. Surgical treatment of Crohn’s disease of the small bowel or ileocecum. World J Surg 1988;12:180e5. McGovern DP, Travis SP. Thiopurine therapy: when to start and when to stop. Eur J Gastroenterol Hepatol 2003;15:219e23.

605

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 327. 328. 329. 330. 331. 332. 333. 334. 335. 336. 337. 338. 339. 340. 341. 342. 343.

344. 345. 346. 347. 348. 349. 350. 351. 352. 353. 354. 355. 356. 357. 358.

606

Clare DF, Alexander FC, Mike S, et al. Accelerated infliximab infusions are safe and well tolerated in patients with inflammatory bowel disease. Eur J Gastroenterol Hepatol 2009;21:71e5. Schwartz DA, Loftus EV Jr, Tremaine WJ, et al. The natural history of fistulizing Crohn’s disease in Olmsted County, Minnesota. Gastroenterology 2002;122:875e80. Hellers G, Bergstrand O, Ewerth S, et al. Occurrence and outcome after primary treatment of anal fistulae in Crohn’s disease. Gut 1980;21:525e7. Buchmann P, Keighley MR, Allan RN, et al. Natural history of perianal Crohn’s disease. Ten year follow-up: a plea for conservatism. Am J Surg 1980;140:642e4. Rutgeerts P. Review article: treatment of perianal fistulizing Crohn’s disease. Aliment Pharmacol Ther 2004;(20 Suppl 4):106e10. Nordgren S, Fasth S, Hulten L. Anal fistulas in Crohn’s disease: incidence and outcome of surgical treatment. Int J Colorectal Dis 1992;7:214e18. Sandborn WJ, Fazio VW, Feagan BG, et al. AGA technical review on perianal Crohn’s disease. Gastroenterology 2003;125:1508e30. Vermeire S, Van Assche G, Rutgeerts P. Perianal Crohn’s disease: classification and clinical evaluation. Dig Liver Dis 2007;39:959e62. Gaertner WB, Decanini A, Mellgren A, et al. Does infliximab infusion impact results of operative treatment for Crohn’s perianal fistulas? Dis Colon Rectum 2007;50:1754e60. Ardizzone S, Maconi G, Colombo E, et al. Perianal fistulae following infliximab treatment: clinical and endosonographic outcome. Inflamm Bowel Dis 2004;10:91e6. Bell SJ, Halligan S, Windsor AC, et al. Response of fistulating Crohn’s disease to infliximab treatment assessed by magnetic resonance imaging. Aliment Pharmacol Ther 2003;17:387e93. Bernstein LH, Frank MS, Brandt LJ, et al. Healing of perineal Crohn’s disease with metronidazole. Gastroenterology 1980;79:357e65. Lichtenstein GR, Yan S, Bala M, et al. Infliximab maintenance treatment reduces hospitalizations, surgeries, and procedures in fistulizing Crohn’s disease. Gastroenterology 2005;128:862e9. Sands BE, Blank MA, Patel K, et al. Long-term treatment of rectovaginal fistulas in Crohn’s disease: response to infliximab in the ACCENT II Study. Clin Gastroenterol Hepatol 2004;2:912e20. American Gastroenterological Association medical position statement: perianal Crohn’s disease. Gastroenterology 2003;125:1503e7. Talbot C, Sagar PM, Johnston MJ, et al. Infliximab in the surgical management of complex fistulating anal Crohn’s disease. Colorectal Dis 2005;7:164e8. Topstad DR, Panaccione R, Heine JA, et al. Combined seton placement, infliximab infusion, and maintenance immunosuppressives improve healing rate in fistulizing anorectal Crohn’s disease: a single center experience. Dis Colon Rectum 2003;46:577e83. O’Connor L, Champagne BJ, Ferguson MA, et al. Efficacy of anal fistula plug in closure of Crohn’s anorectal fistulas. Dis Colon Rectum 2006;49:1569e73. Jess T, Winther KV, Munkholm P, et al. Mortality and causes of death in Crohn’s disease: follow-up of a population-based cohort in Copenhagen County, Denmark. Gastroenterology 2002;122:1808e14. Kwan LY, Conklin JL, Papadakis KA. Esophageal Crohn’s disease treated successfully with adalimumab. Inflamm Bowel Dis 2007;13:639e40. Hassan C, Zullo A, De Francesco V, et al. Systematic review: Endoscopic dilatation in Crohn’s disease. Aliment Pharmacol Ther 2007;26:1457e64. East JE, Brooker JC, Rutter MD, et al. A pilot study of intrastricture steroid versus placebo injection after balloon dilatation of Crohn’s strictures. Clin Gastroenterol Hepatol 2007;5:1065e9. Bernell O, Lapidus A, Hellers G. Risk factors for surgery and postoperative recurrence in Crohn’s disease. Ann Surg 2000;231:38e45. Cosnes J, Nion-Larmurier I, Beaugerie L, et al. Impact of the increasing use of immunosuppressants in Crohn’s disease on the need for intestinal surgery. Gut 2005;54:237e41. Vermeire S, van Assche G, Rutgeerts P. Review article: Altering the natural history of Crohn’s diseaseeevidence for and against current therapies. Aliment Pharmacol Ther 2007;25:3e12. Rutgeerts P, Geboes K, Vantrappen G, et al. Predictability of the postoperative course of Crohn’s disease. Gastroenterology 1990;99:956e63. Kim NK, Senagore AJ, Luchtefeld MA, et al. Long-term outcome after ileocecal resection for Crohn’s disease. Am Surg 1997;63:627e33. Nordgren SR, Fasth SB, Oresland TO, et al. Long-term follow-up in Crohn’s disease. Mortality, morbidity, and functional status. Scand J Gastroenterol 1994;29:1122e8. Marchal L, D’Haens G, Van Assche G, et al. The risk of post-operative complications associated with infliximab therapy for Crohn’s disease: a controlled cohort study. Aliment Pharmacol Ther 2004;19:749e54. Young-Fadok TM, HallLong K, McConnell EJ, et al. Advantages of laparoscopic resection for ileocolic Crohn’s disease. Improved outcomes and reduced costs. Surg Endosc 2001;15:450e4. Tan JJ, Tjandra JJ. Laparoscopic surgery for Crohn’s disease: a meta-analysis. Dis Colon Rectum 2007;50:576e85. Eshuis EJ, Polle SW, Slors JF, et al. Long-term surgical recurrence, morbidity, quality of life, and body image of laparoscopic-assisted vs. open ileocolic resection for Crohn’s disease: a comparative study. Dis Colon Rectum 2008;51:858e67.

359. 360. 361. 362. 363. 364. 365. 366. 367. 368. 369. 370. 371. 372. 373. 374. 375. 376. 377. 378. 379. 380. 381.

382. 383. 384. 385. 386. 387. 388. 389.

Lawes DA, Motson RW. Avoidance of laparotomy for recurrent disease is a longterm benefit of laparoscopic resection for Crohn’s disease. Br J Surg 2006;93:607e8. Stocchi L, Milsom JW, Fazio VW. Long-term outcomes of laparoscopic versus open ileocolic resection for Crohn’s disease: follow-up of a prospective randomized trial. Surgery 2008;144:622e7; discussion 627e8. Polle SW, Slors JF, Weverling GJ, et al. Recurrence after segmental resection for colonic Crohn’s disease. Br J Surg 2005;92:1143e9. Tekkis PP, Purkayastha S, Lanitis S, et al. A comparison of segmental vs subtotal/ total colectomy for colonic Crohn’s disease: a meta-analysis. Colorectal Dis 2006;8:82e90. Fichera A, McCormack R, Rubin MA, et al. Long-term outcome of surgically treated Crohn’s colitis: a prospective study. Dis Colon Rectum 2005;48:963e9. Fearnhead NS, Chowdhury R, Box B, et al. Long-term follow-up of strictureplasty for Crohn’s disease. Br J Surg 2006;93:475e82. Michelassi F, Upadhyay GA. Side-to-side isoperistaltic strictureplasty in the treatment of extensive Crohn’s disease. J Surg Res 2004;117:71e8. Yamamoto T, Fazio VW, Tekkis PP. Safety and efficacy of strictureplasty for Crohn’s disease: a systematic review and meta-analysis. Dis Colon Rectum 2007;50:1968e86. Heuman R, Boeryd B, Bolin T, et al. The influence of disease at the margin of resection on the outcome of Crohn’s disease. Br J Surg 1983;70:519e21. Becker JM. Surgical therapy for ulcerative colitis and Crohn’s disease. Gastroenterol Clin North Am 1999;28:371e90, viiieix. Onali S, Petruzziello C, Calabrese E, et al. Frequency, pattern, and risk factors of postoperative recurrence of Crohn’s disease after resection different from ileocolonic. J Gastrointest Surg 2009;13:246e52. Ikeuchi H, Kusunoki M, Yamamura T. Long-term results of stapled and hand-sewn anastomoses in patients with Crohn’s disease. Dig Surg 2000;17:493e6. Munoz-Juarez M, Yamamoto T, Wolff BG, et al. Wide-lumen stapled anastomosis vs. conventional end-to-end anastomosis in the treatment of Crohn’s disease. Dis Colon Rectum 2001;44:20e5; discussion 25e6. Simillis C, Purkayastha S, Yamamoto T, et al. A meta-analysis comparing conventional end-to-end anastomosis vs. other anastomotic configurations after resection in Crohn’s disease. Dis Colon Rectum 2007;50:1674e87. McLeod RS, Wolff BG, Ross S, et al. Recurrence of Crohn’s disease after ileocolic resection is not affected by anastomotic type: results of a multicenter, randomized, controlled trial. Dis Colon Rectum 2009;52:919e27. Blum E, Katz JA. Postoperative therapy for Crohn’s disease. Inflamm Bowel Dis 2009;15:463e72. Hanauer SB, Korelitz BI, Rutgeerts P, et al. Postoperative maintenance of Crohn’s disease remission with 6-mercaptopurine, mesalamine, or placebo: a 2-year trial. Gastroenterology 2004;127:723e9. Ardizzone S, Maconi G, Sampietro GM, et al. Azathioprine and mesalamine for prevention of relapse after conservative surgery for Crohn’s disease. Gastroenterology 2004;127:730e40. Peyrin-Biroulet L, Deltenre P, Ardizzone S, et al. Azathioprine and 6mercaptopurine for the prevention of postoperative recurrence in Crohn’s disease: a meta-analysis. Am J Gastroenterol 2009;104:2089e96. Duerksen DR, Fallows G, Bernstein CN. Vitamin B12 malabsorption in patients with limited ileal resection. Nutrition 2006;22:1210e13. Behrend C, Jeppesen PB, Mortensen PB. Vitamin B12 absorption after ileorectal anastomosis for Crohn’s disease: effect of ileal resection and time span after surgery. Eur J Gastroenterol Hepatol 1995;7:397e400. Suchy FS, Balistreri WF. Ileal dysfunction in Crohn’s disease assessed by the postprandial serum bile acid response. Gut 1981;22:948e52. Farkkila MA, Kairemo KJ, Taavitsainen MJ, et al. Plasma lathosterol as a screening test for bile acid malabsorption due to ileal resection: correlation with 75SeHCAT test and faecal bile acid excretion. Clin Sci (Lond) 1996;90:315e19. Brydon WG, Nyhlin H, Eastwood MA, et al. Serum 7 alpha-hydroxy-4-cholesten-3one and selenohomocholyltaurine (SeHCAT) whole body retention in the assessment of bile acid induced diarrhoea. Eur J Gastroenterol Hepatol 1996;8:117e23. Minderhoud IM, Oldenburg B, van Dam PS, et al. High prevalence of fatigue in quiescent inflammatory bowel disease is not related to adrenocortical insufficiency. Am J Gastroenterol 2003;98:1088e93. Eaden JA, Mayberry JF. Guidelines for screening and surveillance of asymptomatic colorectal cancer in patients with inflammatory bowel disease. Gut 2002;(51 Suppl 5): V10e12. Soetikno RM, Lin OS, Heidenreich PA, et al. Increased risk of colorectal neoplasia in patients with primary sclerosing cholangitis and ulcerative colitis: a metaanalysis. Gastrointest Endosc 2002;56:48e54. Ullman T, Croog V, Harpaz N, et al. Progression of flat low-grade dysplasia to advanced neoplasia in patients with ulcerative colitis. Gastroenterology 2003;125:1311e19. Collins PD, Mpofu C, Watson AJ, et al. Strategies for detecting colon cancer and/ or dysplasia in patients with inflammatory bowel disease. Cochrane Database Syst Rev 2006;(2):CD000279. Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002). Gut 2010;59:666e89. Cornish J, Tan E, Teare J, et al. A meta-analysis on the influence of inflammatory bowel disease on pregnancy. Gut 2007;56:830e7.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Guidelines 390. 391. 392. 393. 394. 395. 396. 397. 398. 399. 400. 401. 402.

403. 404. 405. 406. 407. 408. 409. 410. 411.

Bar-Oz B, Bulkowstein M, Benyamini L, et al. Use of antibiotic and analgesic drugs during lactation. Drug Saf 2003;26:925e35. Park-Wyllie L, Mazzotta P, Pastuszak A, et al. Birth defects after maternal exposure to corticosteroids: prospective cohort study and meta-analysis of epidemiological studies. Teratology 2000;62:385e92. Norgard B, Hundborg HH, Jacobsen BA, et al. Disease activity in pregnant women with Crohn’s disease and birth outcomes: a regional Danish cohort study. Am J Gastroenterol 2007;102:1947e54. Gisbert JP. Safety of immunomodulators and biologics for the treatment of inflammatory bowel disease during pregnancy and breast-feeding. Inflamm Bowel Dis;16:881e95. Branche J, Cortot A, Bourreille A, et al. Cyclosporine treatment of steroidrefractory ulcerative colitis during pregnancy. Inflamm Bowel Dis 2009;15:1044e8. Katz JA, Antoni C, Keenan GF, et al. Outcome of pregnancy in women receiving infliximab for the treatment of Crohn’s disease and rheumatoid arthritis. Am J Gastroenterol 2004;99:2385e92. Mahadevan U. Gastrointestinal medications in pregnancy. Best Pract Res Clin Gastroenterol 2007;21:849e77. Kane S, Ford J, Cohen R, et al. Absence of infliximab in infants and breast milk from nursing mothers receiving therapy for Crohn’s disease before and after delivery. J Clin Gastroenterol 2009;43:613e16. Williams H, Walker D, Orchard TR. Extraintestinal manifestations of inflammatory bowel disease. Curr Gastroenterol Rep 2008;10:597e605. LaRusso NF, Shneider BL, Black D, et al. Primary sclerosing cholangitis: summary of a workshop. Hepatology 2006;44:746e64. Mendes FD, Levy C, Enders FB, et al. Abnormal hepatic biochemistries in patients with inflammatory bowel disease. Am J Gastroenterol 2007;102:344e50. Khan SA, Davidson BR, Goldin R, et al. Guidelines for the diagnosis and treatment of cholangiocarcinoma: consensus document. Gut 2002;(51 Suppl 6):VI1e9. Jess T, Loftus EV Jr, Velayos FS, et al. Risk factors for colorectal neoplasia in inflammatory bowel disease: a nested case-control study from Copenhagen county, Denmark and Olmsted county, Minnesota. Am J Gastroenterol 2007;102:829e36. Broome U, Lofberg R, Veress B, et al. Primary sclerosing cholangitis and ulcerative colitis: evidence for increased neoplastic potential. Hepatology 1995;22:1404e8. Cullen SN, Rust C, Fleming K, et al. High dose ursodeoxycholic acid for the treatment of primary sclerosing cholangitis is safe and effective. J Hepatol 2008;48:792e800. Lindor KD, Kowdley KV, Luketic VA, et al. High-dose ursodeoxycholic acid for the treatment of primary sclerosing cholangitis. Hepatology 2009;50:808e14. Chapman RW. Primary sclerosing cholangitis: what is the role of ursodeoxycholic acid in therapy for PSC? Nat Rev Gastroenterol Hepatol 2010;7:74e5. Pardi DS, Loftus EV Jr, Kremers WK, et al. Ursodeoxycholic acid as a chemopreventive agent in patients with ulcerative colitis and primary sclerosing cholangitis. Gastroenterology 2003;124:889e93. Tung BY, Emond MJ, Haggitt RC, et al. Ursodiol use is associated with lower prevalence of colonic neoplasia in patients with ulcerative colitis and primary sclerosing cholangitis. Ann Intern Med 2001;134:89e95. Gasche C, Lomer MC, Cavill I, et al. Iron, anaemia, and inflammatory bowel diseases. Gut 2004;53:1190e7. Gasche C, Berstad A, Befrits R, et al. Guidelines on the diagnosis and management of iron deficiency and anemia in inflammatory bowel diseases. Inflamm Bowel Dis 2007;13:1545e53. Bermejo F, Garcia-Lopez S. A guide to diagnosis of iron deficiency and iron deficiency anemia in digestive diseases. World J Gastroenterol 2009;15:4638e43.

Gut 2011;60:571e607. doi:10.1136/gut.2010.224154

412. 413.

414. 415.

416. 417. 418. 419. 420. 421. 422. 423. 424. 425. 426. 427. 428. 429. 430.

Vijverman A, Piront P, Belaiche J, et al. Evolution of the prevalence and characteristics of anemia in inflammatory bowel diseases between 1993 and 2003. Acta Gastroenterol Belg 2006;69:1e4. Kulnigg S, Stoinov S, Simanenkov V, et al. A novel intravenous iron formulation for treatment of anemia in inflammatory bowel disease: the ferric carboxymaltose (FERINJECT) randomized controlled trial. Am J Gastroenterol 2008; 103:1182e92. Erichsen K, Ulvik RJ, Nysaeter G, et al. Oral ferrous fumarate or intravenous iron sucrose for patients with inflammatory bowel disease. Scand J Gastroenterol 2005;40:1058e65. Schroder O, Mickisch O, Seidler U, et al. Intravenous iron sucrose versus oral iron supplementation for the treatment of iron deficiency anemia in patients with inflammatory bowel diseaseea randomized, controlled, open-label, multicenter study. Am J Gastroenterol 2005;100:2503e9. Schreiber S, Howaldt S, Schnoor M, et al. Recombinant erythropoietin for the treatment of anemia in inflammatory bowel disease. N Engl J Med 1996;334:619e23. Rahier JF, Yazdanpanah Y, Colombel JF, et al. The European (ECCO) Consensus on infection in IBD: what does it change for the clinician? Gut 2009;58:1313e15. Salisbury D, Ramsay M, Noakes K. Immunisation Against Infectious Disease. Joint Committee on Vaccination and Immunisation, Great Britain Dept. of Health London: Stationery Office, Dept of Health, London, 2006. Walker JR, Ediger JP, Graff LA, et al. The Manitoba IBD cohort study: a populationbased study of the prevalence of lifetime and 12-month anxiety and mood disorders. Am J Gastroenterol 2008;103:1989e97. Fuller-Thomson E, Sulman J. Depression and inflammatory bowel disease: findings from two nationally representative Canadian surveys. Inflamm Bowel Dis 2006;12:697e707. Graff LA, Walker JR, Bernstein CN. Depression and anxiety in inflammatory bowel disease: a review of comorbidity and management. Inflamm Bowel Dis 2009;15:1105e18. Graff LA, Walker JR, Clara I, et al. Stress coping, distress, and health perceptions in inflammatory bowel disease and community controls. Am J Gastroenterol 2009;104:2959e69. Santos J, Alonso C, Vicario M, et al. Neuropharmacology of stress-induced mucosal inflammation: implications for inflammatory bowel disease and irritable bowel syndrome. Curr Mol Med 2008;8:258e73. Mawdsley JE, Rampton DS. Psychological stress in IBD: new insights into pathogenic and therapeutic implications. Gut 2005;54:1481e91. Maunder RG, Levenstein S. The role of stress in the development and clinical course of inflammatory bowel disease: epidemiological evidence. Curr Mol Med 2008;8:247e52. Bitton A, Dobkin PL, Edwardes MD, et al. Predicting relapse in Crohn’s disease: a biopsychosocial model. Gut 2008;57:1386e92. von Wietersheim J, Kessler H. Psychotherapy with chronic inflammatory bowel disease patients: a review. Inflamm Bowel Dis 2006;12:1175e84. Mikocka-Walus AA, Turnbull DA, Moulding NT, et al. Antidepressants and inflammatory bowel disease: a systematic review. Clin Pract Epidemiol Ment Health 2006;2:24. Blum RW, Garell D, Hodgman CH, et al. Transition from child-centered to adult health-care systems for adolescents with chronic conditions. A position paper of the Society for Adolescent Medicine. J Adolesc Health 1993;14:570e6. Viner R. Transition from paediatric to adult care. Bridging the gaps or passing the buck? Arch Dis Child 1999;81:271e5.

607

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus

Roles of Kru¨ppel-like factor 4 in oesophageal epithelial cells in Barrett’s epithelium development Hideaki Kazumori,1 Shunji Ishihara,1 Yoshiko Takahashi,1 Yuji Amano,2 Yoshikazu Kinoshita1 < An additional figure is

published online only. To view this file please visit the journal online (http://gut.bmj.com). 1

Second Department of Internal Medicine, Shimane University, School of Medicine, Japan 2 Division of Gastrointestinal Endoscopy, Shimane University Hospital, 89-1 Enya-cho, Izumo, Shimane 693-8501, Japan Correspondence to Dr Hideaki Kazumori, Second Department of Internal Medicine, Shimane University, School of Medicine, 89-1 Enya-cho, Izumo, Shimane 693-8501, Japan; [email protected] Revised 9 November 2010 Accepted 11 November 2010 Published Online First 30 December 2010

ABSTRACT Objectives The mechanism of transformation to intestinal metaplasia in Barrett’s oesophagus has not been clarified. We previously reported that bile acids activate the Cdx2 promoter via nuclear factor kappa B (NF-kB) and stimulate production of Cdx2 protein in oesophageal keratinocytes, resulting in production of intestinal-type mucin. Kru¨ppel-like factor 4 (KLF4) is an important transcription factor in the development of intestinal mucosa and has similar functions as Cdx2. In the present study, we investigated the direct effects of bile acids on KLF4 expression as well as the precise mechanisms of expression in cultured oesophageal squamous epithelial cells. Methods We investigated the expression of KLF4 in rat and human Barrett’s epithelium specimens, while the response of that expression to bile acids was studied using a KLF4 promoter luciferase assay. In addition, oesophageal squamous epithelial cells were transfected with a KLF4 expression vector, after which their possible transformation into intestinal-type epithelial cells was investigated. Results In both rat and human tissues, Barrett’s epithelium strongly expressed KLF4. Furthermore, a bile acids mixture increased KLF4 promoter activity, and mRNA and protein expression in oesophageal epithelial cells. Results from mutation analysis of the KLF4 promoter suggested that the NF-kB binding site is responsible for bile acid-induced activation of the KLF4 promoter. In addition, KLF4 and Cdx2 stimulated each other by directly binding to the promoter of the other, while transfection of the KLF4 expression vector in oesophageal epithelial cells induced production of MUC2 protein. Conclusion Bile acid-induced sequential expression of KLF4 followed by MUC2 production may have an important role in the development of Barrett’s epithelium.

INTRODUCTION Barrett’s oesophagus is an acquired condition, in which stratiﬁed squamous epithelium is replaced by metaplastic columnar epithelium in the distal oesophagus.1 The condition is associated with chronic gastro-oesophageal reﬂux disease (GORD)2 and reﬂux of duodenal contents with bile acids is generally considered to be one of the most important risk factors in its development.3 In association with tissue damage and regeneration, one or a few stem cells may attempt to adapt to this new environment by altering the patterns of some gene expressions, and thus undergo profound phenotypic changes that lead to a different type of epithelium that is more resistant to such a novel 608

Significance of this study What is already known about this subject?

< Cdx2 is a key mediator in the development of

Barrett’s oesophagus.

< Bile acids directly augment Cdx2 via NF-kB. < Cdx1 is also an important molecular mediator of

Barrett’s oesophagus.

What are the new findings?

< KLF4 is expressed in Barrett’s epithelium. < The expression of KLF4 in oesophageal kerati-

nocytes in response to bile acids induces metaplastic changes during Barrett’s epithelium development. < The transcriptional network related to KLF4 and Cdx2 has important roles in development of this disease.

How might it impact on clinical practice in the foreseeable future? < Molecular targets for treatment of Barrett’s oesophagus will be defined.

environment. The causal link between bile acid reﬂux and alterations of some transcription factors has been studied; however, the precise mechanism of promotion of Barrett’s epithelium formation by bile acid reﬂux remains to be characterised. A number of studies have found that Cdx2 is a key mediator in the development of Barrett’s oesophagus.4 5 We previously reported a two-step mechanism involved in the development of Barrett’s epithelium, in which bile acids activate the Cdx2 promoter via nuclear factor kappa B (NF-kB) and stimulate the production of Cdx2 protein in oesophageal immature keratinocytes, with a resulting production of intestinal type mucin.6 In addition to Cdx2, we also recently showed that bile acids induce the expression of Cdx1 in oesophageal immature keratinocytes, and demonstrated an interplay mechanism between Cdx1 and Cdx2 that causes upregulation of each other by directly binding to the promoter of the other, stimulating the development of Barrett’s epithelium.7 Krüppel-like factors (KLFs) are zinc ﬁngercontaining transcription factors that exhibit homology to the Drosophila melanogaster segmentation gene product Krüppel. KLFs comprise a family of evolutionarily conserved zinc ﬁnger transcription factors that regulate numerous biological processes, including proliferation, differentiation, development Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus and apoptosis.8 Among them, KLF4 (gut-enriched Krüppel-like factor) is highly expressed in epithelial cells of the small and large intestines.9 The expression pattern of KLF4 is similar to that of Cdx2 in the intestine, as it is expressed mainly in non-proliferating differentiating and differentiated cells of the upper crypt and villus/surface mucosa, where Cdx2 is also expressed.8 During embryogenesis, the expression of KLF4 begins to rise, which correlates with a critical period of gut epithelium morphogenesis, similar to that of Cdx2.10 Furthermore, colonic goblet cells in KLF4/ mice do not show a normal goblet cell morphology and the goblet cell marker MUC2 exhibits patchy expression throughout the colonic epithelium.11 These ﬁndings suggest that KLF4 plays a fundamental role in the development of intestinal mucosa. With regard to carcinogenesis, it was reported that KLF4 expression is downregulated in human adenomatous polyps and cancer of the colon.12 Notably, in colorectal adenomas and adenocarcinomas, the level of Cdx2 protein is markedly reduced.13 Also, Cdx2 was shown to activate a KLF4 promoter construct in Chinese hamster ovary (CHO) cells and colon cancer related RKO cells,14 15 while an inter-regulation mechanism between KLF4 and Cdx2 has been speculated. In the present study, we investigated whether alterations of KLF4 expression in response to bile acids in oesophageal keratinocytes induce metaplastic changes during Barrett’s epithelium development. Furthermore, we investigated the transcriptional network connecting KLF4 and Cdx2 in development of the disease.

MATERIALS AND METHODS Rat model of Barrett’s oesophagus To induce Barrett’s oesophagus, we employed Levrat’s model with minor modiﬁcations, as previously described.6 7 16 In brief, the gastro-oesophageal junction was cut and the oesophageal end separated. The distal end of the oesophagus was then reimplanted 2 cm beyond the ligament of Treitz in an end-toside fashion into a loop of the jejunum and the proximal end of the stomach was ligated. Six months after formation of oesophagealejejunal anastomoses, the rats were killed and their oesophagi removed.

Patients and tissues Human oesophageal tissues were collected after obtaining informed, written consent from all subjects. During endoscopy procedures, biopsy specimens of normal squamous mucosa from the distal oesophagus (n¼6) and Barrett’s oesophagus without dysplasia (n¼6) were taken, then snap-frozen in liquid nitrogen. Barrett’s oesophagus was histologically deﬁned as the presence of columnar epithelium containing goblet cell metaplasia.

Cell culture and bile acid treatment Five cell lines, including Het-1A (a human normal oesophageal cell line immortalised by viral SV40 transfection; American Type Culture Collection, ATCC, Manassas, Virginia, USA), OE33 (a human oesophageal adenocarcinoma cell line; European Collection of Cell Cultures, ECACC, Salisbury, Wiltshire, UK), OE19 (a human cell line established from an adenocarcinoma obtained from the gastric cardia/oesophageal gastric junction; ECACC), SW480 (a human colorectal adenocarcinoma cell line, ATCC) and HeLa (a human cervical adenocarcinoma cell line, ATCC), were used in this study. Primary cultures of oesophageal keratinocytes from normal rat oesophagi were established, as previously described.6 Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

A mixture of bile acids (Sigma Chemicals, St. Louis, Missouri, USA), which included cholic acid, glycocholic acid and taurocholic acid, was used as a stimulant, as previously described.7

Vector construction and luciferase assay We ampliﬁed 1700 bp of the KLF4 promoter (accession No. AF117109) by PCR, then cloned that into the MluI and BglII sites of a pGL3-basic luciferase vector (Promega, Madison, Wisconsin, USA) to generate pKLF4/1700-Luc (1735 to 36), which generated pKLF4/1080-Luc (1115 to 36), pKLF4/425Luc (460 to 36), pKLF4/233-Luc (268 to 36), and pKLF4/ 35-Luc (70 to 36). The position +1 refers to the major transcription start site identiﬁed in the KLF4 gene.9 We also ampliﬁed 1541 bp of the Cdx2 promoter (accession No. NC_000071) by PCR, then cloned that into the KpnI and BglII sites of a pGL3-basic luciferase vector to generate pCdx2/ 1541-Luc (1415 to +125), which generated pCdx2/1014-Luc (888 to +125), pCdx2/631-Luc (506 to +125), pCdx2/438Luc (313 to +125), pCdx2/319-Luc (194 to +125), pCdx2/ 219-Luc (94 to +125), and pCdx2/74-Luc (+52 to +125), as previously described.6 Furthermore, 1750 bp of the MUC2 promoter (accession No. AF221746) was ampliﬁed by PCR, then cloned into the MluI and BglII sites of a pGL3-basic luciferase vector to generate pMUC2/1750-Luc (1804 to 55), which generated pMUC2/823-Luc (877 to 55), pMUC2/463-Luc (517 to 55), pMUC2/214-Luc (268 to 55), pMUC2/ 80-Luc (134 to 55), and pMUC2/39-Luc (93 to 55). The position +1 refers to the major transcription start site identiﬁed in the MUC2 gene.17 As an internal control for the dual luciferase assay, pRL-TATA-Renilla-Luc was used.6 To produce mutated KLF4 promoter constructs for pM/KLF4-Luc, 59 -ggcggccgccagtacttcaccggccgagagagcgagcgcggctcc-39 was used (nucleotide substitutions indicated in bold), to produce mutated Cdx2 promoter constructs for pM/Cdx2-Luc, 59 -cggcgggtcattccaagtctctacagcttactggcaaggaggtgggaggaaa-39 was used, and to produce mutated MUC2 promoter constructs, for pM/MUC2Luc, 59 -cttggcaaataatacgtgaatatttcgcacctccctcgtcctccgccctcg-39 was used. cDNA encoding full-length mouse KLF4 (NCBI NM-010637) was ampliﬁed by PCR and cloned into a pcDNA5/FRT/V5-HisTOPO Vector (Invitrogen, Carlsbad, California, USA). Vector DNA without KLF4 sequences was used as a negative control. A Cdx2 expression vector was also constructed, as previously reported.6 Het-1A, OE33, and OE19 cells were separately cultured and transfected with 0.5 mg of each promoter vector and 0.02 mg of pRL-TATA-Renilla-Luc in each well, with Lipofectamine 2000 (Invitrogen). At 24 h after transfection of the luciferase vectors, the cells were stimulated with various concentrations of the bile acids mixture or the vehicle alone for 3 h, then cell lysates were used to determine luciferase activity. Also, Het-1A cells were cultured and transfected with 0.2 mg of each indicated promoter vector and a total of 0.2 mg of each indicated expression vector or an empty vector, along with 0.02 mg of pRL-TATA-Renilla-Luc in each well for 24 h, then the cell lysates were used for measurement of luciferase activity.

Immunohistochemistry Immunohistochemistry was performed as previously described.6 To identify KLF4-expressing cells, tissue sections were incubated with the anti-KLF4 antibody (1:100; Medical & Biological Laboratories, Nagoya, Japan), followed by incubation with secondary biotinylated anti-rabbit immunoglobulin (DAKO, Carpinteria, California, USA). Bound antibodies were detected 609

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus using a 3-amino-9-ethylcarbazole substrateechromogen system (DAKO). The sections were counter-stained with haematoxylin.

Chromatin immunoprecipitation

Extraction of total RNA was performed as previously described.6 DNase I-treated RNA was reverse transcribed into cDNA using a ReverTra Ace a kit (Stratagene Toyobo, Tokyo, Japan). A realtime ﬂuorescence PCR assay based on SYBR Green (Applied Biosystems, Foster City, California, USA) was then performed using the primers described in table 1. Primary cultured cells were transfected with control nonspeciﬁc siRNA (Qiagen, Hilden, Germany), p50 siRNA (Santa Cruz Biotechnology, Santa Cruz, California, USA), or p65 siRNA (Santa Cruz Biotechnology) using Lipofectamine 2000. The reduced levels of p50 or p65 mRNA expression induced by transfection of each siRNA were determined using siRNA speciﬁc primers (Santa Cruz Biotechnology).

Chromatin immunoprecipitation (ChIP) analysis was performed using an EpiQuik Chromatin Immunoprecipitation Kit (Epigentek Group, Brooklyn, New York, USA), according to the method reported by D’Amico et al.18 Het-1A cells were transiently transfected with a KLF4 promoter vector and stimulated with a bile acids mixture for 3 h, after which ChIP analysis was performed. Also, Het-1A cells were transiently transfected with a KLF4 promoter vector, Cdx2 promoter vector, MUC2 promoter vector and KLF4 expression vector, Cdx2 expression vector, or empty vector, after which ChiP analysis was performed. Total DNA prior to immunoprecipitation was used as the input value. Chromatin was immunoprecipitated with anti-KLF4, anti-Cdx2, anti-p50, and anti-p65 antibodies, or IgG as a negative control. Immunoprecipitated DNAeprotein complexes were isolated and a real-time PCR assay was then performed using the primers described in table 1.

Protein extraction and western blot analysis

Statistical analysis

RNA extraction and real-time PCR

Protein extraction and western blot analysis were performed as previously previously.6 The membranes were incubated with anti-KLF4 (1:200; Abnova, Taipei, Taiwan), anti-p50 (1:200; Santa Cruz Biotechnology), anti-p65 (1:200; Santa Cruz Biotechnology), or anti-b-actin (1:3000; Sigma Chemicals) antibodies, followed by horseradish-peroxidase-conjugated antimouse or anti-rabbit immunoglobulin (DAKO).

All data are expressed as the mean 6 SEM. Multiple comparisons were performed with ANOVA, followed by a Dunnett test. Statistical comparisons between two groups were done with a ManneWhitney U test. p values less than 0.05 were considered to be statistically signiﬁcant.

Immunofluorescence cytochemistry

KLF4 mRNA was found to be expressed throughout the gastrointestinal tract of adult rats, with high levels of expression observed in the jejunum, ileum, proximal, and distal colon, while a lower level of expression was found in the oesophagus (ﬁgure 1A).

Immunoﬂuorescence cytochemistry was performed as previously described.6 The cells were labelled with anti-KLF4 (1:100), anti-Cdx2 (1:50; BioGenex, San Ramon, California, USA), antiMUC2 (1:100; Santa Cruz Biotechnology), anti-p50 (1:100), anti-p65 (1:100), and anti-cytokeratin (CK) 20 (1:200; Santa Cruz Biotechnology) antibodies. Binding of the primary antibodies was detected using FITC-conjugated anti-mouse, antirabbit, or anti-goat immunoglobulin, or rhodamine-conjugated anti-mouse or anti-rabbit immunoglobulin (DAKO). The cells were nuclear counter-stained with 49 ,6-diamidino-29 -phenylindole dihydrochloride (DAPI) (Pierce Biotechnology, Rockford, Illinois, USA). Table 1

The sequences of oligonucleotide primers used in this study Sequence (59 to 39 )

Primer Human GAPDH Human KLF4 Human MUC2 Rat GAPDH Rat KLF4 KLF4 promoter, including the NF-kB binding site KLF4 promoter, including GC boxes Cdx2 promoter, including the Sp-1-binding site KLF4 promoter including Cdx2-binding site MUC2 promoter, including the CACCC/Sp-1 element

610

Fw Rv Fw Rv Fw Rv Fw Rv Fw Rv Fw Rv Fw Rv Fw Rv Fw Rv Fw Rv

gaaggtgaaggtcggagtca aatgaaggggtcattgatgg tcccatctttctccacgttc agtcgcttcatgtgggagag ctcccagacaggagaacgag gggatcgcagtggtagttgt gtgaaggtcggtgtgaacg cttgccgtgggtagagtcat caggctgtggcaaaacctat cggtagtgcctggtcagttc gcaagcgagcgagaagttat gagtcctggggactgtgc gtgcgcggagtttgtttatt ccgcgcgcttcttacttat cagccattggtgtctgtgtc ttctttcctcccacctcctt tggccatcggacatacttatc gccccaaagtcaacgaagta tagttcacctggggtgtgtg gacgagggaggtgcgaag

RESULTS Expression of KLF4 mRNA in adult rat tissues

Immunohistochemistry examinations of rat Barrett’s epithelium Six months after the procedure, columnar-lined epithelia consisting of absorptive cells and goblet cells were observed above the oesophagealejejunostomy in the rats. KLF4-positive cells with nuclear staining were observed in the columnar epithelia above the oesophagealejejunostomy, mainly in the surface villi, whereas there was a small number of cells in the crypts (ﬁgure 1B).

Expression of KLF4 mRNA in human normal oesophagus and Barrett’s oesophagus We also determined KLF4 mRNA expression levels in endoscopic biopsy specimens of normal oesophagus and Barrett’s oesophagus obtained from the human subjects. KLF4 mRNA expression levels of Barrett’s oesophagus were signiﬁcantly higher than those of normal squamous epithelium (ﬁgure 1C).

Effects of bile acids on KLF4 promoter activity The bile acids mixture had a stimulatory effect on KLF4 promoter activity in a dose-dependent manner, with an approximately twofold increase in transcriptional activation in Het-1A, OE33, and OE19 cells (ﬁgure 2A,B,C). We constructed a series of reporter plasmids containing different lengths of the KLF4 promoter. The transcriptional activity of these constructs was analysed in Het-1A cells with 200 mm of the bile acids mixture. The plasmid pKLF4/1080-Luc exhibited a similar level of activation by the bile acids mixture as that shown by pKLF4/ 1700-Luc. However, the plasmids pKLF4/425-Luc, pKLF4/ 233-Luc, and pGL3-basic without the KLF4 promoter showed no activation response to bile acid stimulation (ﬁgure 2D). These results revealed that bile acid-induced activation of the KLF4-promoter is controlled by a site located between 1155 Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus Figure 1 (A) Tissue distribution of KLF4 mRNA in gastrointestinal tissues. RNA was extracted from various rat tissues and analysed by real-time PCR for KLF4 expression. Data were normalised to GAPDH mRNA. Results are expressed as the mean 6 SEM of three experiments. (B) Immunohistochemistry findings for KLF4. Barrett’s epithelium was examined 6 months after performance of oesophagealejejunal anastomosis. KLF4-positive cells were mainly observed in surface villi, though a small number of cells were observed in the crypts in Barrett’s epithelium. Normal oesophagus and colon specimens are shown. KLF4-positive cells in the normal colon were localised predominantly on the surface epithelium. In the oesophagus, KLF4-positive cells were localised on the cells of the suprabasal layer, though the level was low. (C) Expression of KLF4 mRNA in human normal oesophagus and Barrett’s oesophagus. Results are expressed as the mean 6 SEM of six samples. *p<0.05 vs. normal oesophagus. and 460. To determine whether bile acids function as a direct transcriptional activator of KLF4, we examined the KLF4 promoter region for the putative NF-kB binding site using the computational program TESS (http://www.cbil.upenn.edu/). We identiﬁed a putative NF-kB binding site from 605 to 596 (ggcagttccc) and speculated that bile acids might bind to the KLF4 promoter in this region. Therefore, to investigate the role of the NF-kB site following bile acid-induced stimulation of KLF4 expression, the element of the putative NF-kB binding site was mutated, which completely abolished bile acid-induced activation of the KLF4 promoter (ﬁgure 2D). To conﬁrm whether bile acids bind to the KLF4 promoter, a ChIP assay was performed using Het-1A cells. Real-time PCR

analysis was performed to amplify the promoter region of KLF4 from 650 to 424 that contains the NF-kB binding site. The amount of transcript in the bile acid-treated samples was signiﬁcantly higher than that in the vehicle-treated samples of DNA immunoprecipitated with the anti-p50 and anti-p65 antibodies (ﬁgure 2E).

Direct effects of bile acids on KLF4 mRNA and protein expressions in oesophageal epithelial cells To determine whether bile acids augment KLF4 mRNA expression, we investigated the direct effect of a bile acids mixture on KLF4 mRNA expression using Het-1A and OE33 cells, and found that the bile acids augmented KLF4 mRNA expression in

Figure 2 Effects of bile acids on transcriptional activation of KLF4 in (A) Het-1A, (B) OE33, and (C) OE19 cells. Twenty-four hours after transfection with the KLF4 promoter vector, cells were stimulated with various concentrations of the bile acids mixture or vehicle alone for 3 h, then cell lysates were used to determine luciferase activity. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. control. (D) Reporter gene analysis of KLF4 promoter deletion and mutation constructs in Het-1A cells. Twenty-four hours after transfection with the indicated KLF4 promoter vectors, cells were stimulated with the bile acids mixture (200 mM) or vehicle alone for 3 h. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. pKLF4/1700-Luc. (E) Chromatin immunoprecipitation assay. At 24 h after transfection with the KLF4 promoter vector, Het-1A cells were treated with the bile acids mixture (200 mM) or vehicle alone for 3 h. Anti-p50 and anti-p65 antibody immunoprecipitated DNA was purified and analysed by real-time PCR for the KLF4 promoter, including the NF-kB binding site. The amount of precipitated DNA was normalised to input DNA. Results are expressed as the mean 6 SEM of three experiments. *p<0.05 vs. control. Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

611

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus a dose-dependent manner (ﬁgure 3A,B). We also investigated the direct effect of the bile acids mixture on KLF4 expression using primary cultured oesophageal keratinocytes, and found that bile acids augmented KLF4 mRNA and protein expression in a dosedependent manner (ﬁgure 3C,D). When examined using immunoﬂuorescence cytochemistry, KLF4 protein expression was augmented with positive nuclear staining by the bile acids mixture (ﬁgure 3E). Furthermore, we investigated the direct effects of the bile acids mixture on p50 and p65 protein expressions. The results of western blotting analysis revealed that the bile acids augmented p50 and p65 protein expressions (ﬁgure 3D). In addition, p50 and p65 nuclear translocations were shown by immunoﬂuorescence cytochemistry following addition of the bile acids mixture (ﬁgure 3E). To determine whether KLF4 induction by bile acids occurs via NF-kB activation, we used an siRNA approach with primary cultured cells. p50 and p65 mRNA expression was signiﬁcantly decreased in p50 siRNA- and p65 siRNA-transfected samples, respectively (data not shown). Furthermore, KLF4 mRNA expression was signiﬁcantly decreased in cells transfected with speciﬁc p50 and p65 siRNAs as compared to the control nonspeciﬁc siRNA transfected cells following bile acid treatment (ﬁgure 3F).

Homologous auto-regulations of KLF4 The speciﬁcity of the KLF4 expression vector was conﬁrmed by western blot analysis (supplementary ﬁgure 1). Transfection of the KLF4 expression vector into Het-1A cells increased KLF4 promoter activity in a dose-dependent manner (ﬁgure 4A). We also constructed a series of reporter plasmids containing different lengths of the KLF4 promoter. The plasmid pKLF4/ 1080-Luc exhibited a level of activation following stimulation with the KLF4 expression vector similar to the activation shown by pKLF4/1700-Luc. However, the plasmids pKLF4/425-Luc and pKLF4/233-Luc, as well as PGL3-basic without the KLF4 promoter showed no activation response to KLF4 stimulation (ﬁgure 4B). These results revealed that KLF4-induced activation of the KLF4-promoter is controlled by a site located between 1115 and 460. A previous report indicated that the KLF4 promoter has three GC boxes that bind to KLF4.14 Therefore, to conﬁrm whether KLF4 binds to the KLF4 promoter, a ChIP assay was performed using Het-1A cells. Real-time PCR analysis was performed to amplify the promoter region of KLF4 from 781 to 614, which contains the three GC boxes. The amount of transcript from the KLF4 transfected cells was signiﬁcantly greater than that from the empty vector transfected cells (ﬁgure 4C).

Figure 3 Effects of bile acids on KLF4 mRNA expression in (A) Het-1A, (B) OE33, and (C) primary cultured cells. Cells were stimulated with various concentrations of the bile acids mixture or vehicle alone for 3 h, then RNA was extracted and subjected to real-time PCR for KLF4. Data were normalised to GAPDH mRNA. Results are expressed as the mean 6 SEM of four experiments. **p<0.01 vs. control. *p<0.05 vs. control. (D) Effects of bile acids on KLF4, p50, and p65 protein expressions in primary cultured cells. Cells were stimulated with various concentrations of the bile acids mixture or vehicle alone for 6 h then protein was extracted and subjected to western blot analysis for KLF4, p50, p65 and b-actin. Blots shown are representative of three separate experiments. (E) Effects of bile acids on KLF4, p50 and p65 protein expressions in primary cultured cells determined by immunofluorescence cytochemistry. After incubation with 200 mM of the bile acids mixture for 6 h, KLF4-, p50- and p65-positive cells with nuclear staining were observed after nuclear counter-staining with DAPI. Results shown are representative of three separate experiments. (F) RNA interference was performed using p50 or p65 siRNA, or control non-specific siRNA. At 48 h after transfection with each siRNA, primary cultured cells were stimulated with the bile acids mixture (200 mM) or vehicle alone for 3 h, then RNA was extracted and subjected to real-time PCR for KLF4. Data were normalised to GAPDH. Data are expressed as the n-fold increase in transcript in the bile acid-stimulated samples over that in the vehicle-treated samples. Results are expressed as the mean 6 SEM of three experiments. *p<0.05 vs. control siRNA transfected samples with bile acids treatment. DAPI, 49 ,6-diamidino-29 -phenylindole dihydrochloride; GAPDH, glyceraldehyde-3-phosphate dehydrogenase. 612

Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus

Figure 4 Homologous auto-regulation mechanism of KLF4. (A) Cells were co-transfected with a KLF4 promoter vector and the indicated amounts of a KLF4 expression or empty vector, then cell lysates were used to determine luciferase activity at 24 h after transfection. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. control. (B) Reporter gene analysis of KLF4 promoter deletion constructs in Het-1A cells. Cells were co-transfected with the indicated KLF4 promoter vectors and a KLF4 expression or empty vector. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. pKLF4/1700-Luc. (C) Chromatin immunoprecipitation assay. Het-1A cells were co-transfected with a KLF4 promoter vector and a KLF4 expression or empty vector for 24 h. Anti-KLF4 antibody immunoprecipitated DNA was purified and analysed by real-time PCR for the KLF4 promoter including GC boxes. The amount of precipitated DNA was normalised to input DNA. Results are expressed as the mean 6 SEM of three experiments. **p<0.01 vs. empty vector.

Heterologous inter-regulation mechanism of Cdx2 stimulated by KLF4 Transfection of the KLF4 expression vector into Het-1A cells increased Cdx2 promoter activity in a dose-dependent manner

(ﬁgure 5A). We constructed a series of reporter plasmids containing different lengths of the Cdx2 promoter. The plasmids pCdx2/1014-Luc, pCdx2/631-Luc, pCdx2/438-Luc, pCdx2/319Luc, and pCdx2/219-Luc exhibited a level of activation following

Figure 5 Heterologous inter-regulation mechanism of Cdx2 stimulated by KLF4. (A) Het-1A cells were cotransfected with a Cdx2 promoter vector and a KLF4 expression or empty vector, then cell lysates were used to determine luciferase activity 24 h after transfection. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. control. (B) Reporter gene analysis of Cdx2 promoter deletion and mutation constructs in Het-1A cells. Cells were co-transfected with the indicated Cdx2 promoter vectors and a KLF4 expression or empty vector. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. pCdx2/1541-Luc. (C) Chromatin immunoprecipitation assay. Het-1A cells were co-transfected with a Cdx2 promoter vector and a KLF4 expression or empty vector for 24 h. Anti-KLF4 antibody immunoprecipitated DNA was purified and analysed by real-time PCR for the Cdx2 promoter including the Sp-1-binding site. The amounts of precipitated DNA were normalised to input DNA. Results are expressed as the mean 6 SEM of three experiments. *p<0.05 vs. empty vector. (D) Immunofluorescence cytochemistry examination for KLF4 and Cdx2 conducted 48 h after transfection of a KLF4 expression vector. Cells transfected with the KLF4 expression vector expressed both KLF4 and Cdx2 proteins after nuclear counter-staining with DAPI. Negative control, cells transfected with an empty vector; positive control, cells transfected with a Cdx2 expression vector. Results shown are representative of three separate experiments. DAPI, 49 ,6-diamidino-29 -phenylindole dihydrochloride. Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

613

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus stimulation with a KLF4 expression vector that was similar to the activation shown by pCdx2/1541-Luc. However, the plasmids pCdx2/74-Luc and PGL3-basic without the Cdx2 promoter showed no activation responses to KLF4 stimulation (ﬁgure 5B). These results revealed that KLF4-induced activation of the KLF4promoter is controlled by a site located between 94 and +52. We identiﬁed a putative Sp-1 binding site from 91 to 82 (tccccgcctct) and speculated that KLF4 might bind to a Cdx2 promoter in this region. Therefore, to investigate the role of the Sp-1 site in KLF4-induced stimulation of Cdx2 expression, that element of the putative Sp-1-binding site was mutated, which completely abolished the KLF4-induced activation of the Cdx2 promoter (ﬁgure 5B). To conﬁrm whether KLF4 binds to the Cdx2 promoter, a ChIP assay was performed using Het-1A cells. Real-time PCR analysis was performed to amplify the region of the Cdx2 promoter from 194 to 48 that contains the Sp-1 binding site. The amount of transcript from the cells transfected with the KLF4 expression vector was signiﬁcantly higher than that from cells transfected with an empty vector (ﬁgure 5C). Finally, Cdx2 protein expression following stimulation with a KLF4 expression vector was evaluated in Het-1A cells using immunoﬂuorescence cytochemistry. Cells transfected with the

KLF4 construct were found to express the Cdx2 transcript (ﬁgure 5D).

Heterologous inter-regulation mechanism of KLF4 stimulated by Cdx2 Transfection of the Cdx2 expression vector into Het-1A cells increased KLF4 promoter activity in a dose-dependent manner (ﬁgure 6A). We also constructed a series of reporter plasmids containing different lengths of the KLF4 promoter. The plasmids pKLF4/1080-Luc, pKLF4/425-Luc, and pKLF4/233-Luc exhibited a level of activation following stimulation with the Cdx2 expression vector similar to the activation shown by pKLF4/ 1541-Luc. However, the plasmids pKLF4/35-Luc and PGL3-basic without the KLF4 promoter showed no activation response to Cdx2 stimulation (ﬁgure 6B). These results revealed that Cdx2induced activation of the KLF4-promoter is controlled by a site located between 268 and 70. We identiﬁed multiple putative Cdx2 binding sites and concluded that Cdx2 might bind to the KLF4 promoter in these regions. To conﬁrm whether Cdx2 binds to the KLF4 promoter, a ChIP assay was performed with Het-1A cells. Real-time PCR analysis was performed to amplify the region of the KLF4 promoter from 259 to 56, which contains multiple Cdx2 binding sites. The amount of transcript from the

Figure 6 Heterologous inter-regulation mechanism of KLF4 stimulated with Cdx2. (A) Het-1A cells were cotransfected with a KLF4 promoter vector and a Cdx2 expression or empty vector, then cell lysates were used to determine luciferase activity at 24 h after transfection. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. control. (B) Reporter gene analysis of KLF4 promoter deletion constructs in Het-1A cells. Cells were co-transfected with the indicated KLF4 promoter vectors and a Cdx2 expression or empty vector. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. pKLF4/1700-Luc. (C) Chromatin immunoprecipitation assay. Het-1A cells were transiently co-transfected with a KLF4 promoter vector and a Cdx2 expression or empty vector for 24 h. Anti-Cdx2 antibody immunoprecipitated DNA was purified and analysed by real-time PCR for the KLF4 promoter including the Cdx2-binding site. The amounts of precipitated DNA were normalised to input DNA. Results are expressed as the mean 6 SEM of three experiments. **p<0.01 vs. empty vector. (D) Forty-eight hours after transfection with a Cdx2 expression vector, an immunofluorescence cytochemistry examination for Cdx2 and KLF4 was conducted. Cells transfected with the Cdx2 expression vector expressed both Cdx2 and KLF4 proteins after nuclear counter-staining with DAPI. Negative control, cells transfected with an empty vector; positive control, cells transfected with a KLF4 expression vector. Results shown are representative of three separate experiments. DAPI, 49 ,6-diamidino-29 -phenylindole dihydrochloride. 614

Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus cells transfected with the Cdx2 expression vector was signiﬁcantly higher than that from cells transfected with an empty vector (ﬁgure 6C). Next, KLF4 protein expression following stimulation with a Cdx2 expression vector was evaluated in Het-1A cells using immunoﬂuorescence cytochemistry and cells transfected with the Cdx2 construct were shown to express the KLF4 transcript (ﬁgure 6D).

Effects of Cdx2 or KLF4 over-expression on oesophageal epithelial cells Transfection of the Cdx2 expression vector into Het-1A cells increased MUC2 promoter activity in a dose-dependent manner (ﬁgure 7A). Transfection of the KLF4 expression vector into Het-1A cells increased MUC2 promoter activity in a dosedependent manner (ﬁgure 7B). We also constructed a series of reporter plasmids containing different lengths of the MUC2 promoter. The plasmids pMUC2/823-Luc, pMUC2/463-Luc, pMUC2/214-Luc, and MUC2/80-Luc exhibited a level of activation following stimulation with the KLF4 expression vector that was similar to the activation shown by pMUC2/1750-Luc. However, the plasmids pMUC2/39-Luc and PGL3-basic without the MUC2 promoter showed no activation response to KLF4 stimulation (ﬁgure 7C). These results revealed that KLF4-induced activation of the MUC2-promoter is controlled by a site located between 134 and 93. We identiﬁed a putative CACCC/Sp-1 binding site from 113 to 101 (gccccacccaccc) and speculated

that KLF4 might bind to the MUC2 promoter in this region. Therefore, to investigate the role of the CACCC/Sp-1 site in KLF4-induced stimulation of MUC2 expression, the element of the putative CACCC/Sp-1-binding site was mutated, which completely abolished KLF4-induced activation of the MUC2 promoter (ﬁgure 7C). To conﬁrm whether KLF4 binds to the MUC2 promoter, a ChIP assay was performed with Het-1A cells. Real-time PCR analysis was performed to amplify the region of the MUC2 promoter from 304 to 84 that contains the CACCC/Sp-1 binding site. The amount of transcript from the cells transfected with the KLF4 expression vector was signiﬁcantly higher than that from the cells transfected with an empty vector (ﬁgure 7D). We also transfected a KLF4 expression vector into Het-1A cells and observed the expression of intestine speciﬁc MUC2 mRNA in those cells. After 48 h, Het-1A cells transfected with the KLF4 expression construct induced MUC2 mRNA expression (ﬁgure 7E). Next, MUC2 protein expression following stimulation with a KLF4 expression vector was evaluated in Het-1A cells using immunoﬂuorescence cytochemistry and those transfected with the KLF4 construct were found to express the MUC2 transcript (ﬁgure 8A). In addition, cells transfected with the KLF4 expression vector induced CK20 expression (ﬁgure 8B).

DISCUSSION The results of the present experiments suggest that KLF4 is an important molecular mediator in the development of Barrett’s epithelium. This is the ﬁrst known study to investigate the role

Figure 7 Effects of over-expression of (A) Cdx2 and (B) KLF4 on MUC2 expression. Het-1Acells were co-transfected with a MUC2 promoter vector and the indicated expression or empty vector, then cell lysates were used to determine luciferase activity 24 h after transfection. Results are expressed as the mean 6 SEM of four experiments. **p<0.01 vs. control. *p<0.05 vs. control. (C) Reporter gene analysis of MUC2 promoter deletion and mutation constructs in Het-1A cells. Cells were co-transfected with the indicated MUC2 promoter vectors and a KLF4 expression or empty vector. Results are expressed as the mean 6 SEM of four experiments. *p<0.05 vs. pMUC2/1750-Luc. (D) Chromatin immunoprecipitation assay. Het-1A cells were co-transfected with a MUC2 promoter vector and a KLF4 expression or empty vector for 24 h. Anti-KLF4 antibody immunoprecipitated DNA was purified and analysed by real-time PCR for the MUC2 promoter including the CACCC/Sp-1 element. The amount of precipitated DNA was normalised to input DNA. Results are expressed as the mean 6 SEM of three experiments. **p<0.01 vs. empty vector. (E) Forty-eight hours after transfection of a KLF4 expression or empty vector, RNA was extracted and subjected to real-time PCR for MUC2. Data were normalised to GAPDH mRNA. Results are expressed as the mean 6 SEM of four experiments. **p<0.01 vs. empty vector. GAPDH, glyceraldehyde-3-phosphate dehydrogenase. Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

615

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus

Figure 8 Effects of over-expression of KLF4 on MUC2 and CK20 expressions. (A) Forty-eight hours after transfection with a KLF4 expression vector, an immunofluorescence cytochemistry examination for KLF4 and MUC2 was conducted. Cells transfected with the KLF4 expression vector expressed both KLF4 and MUC2 proteins. Negative control, cells transfected with an empty vector; positive control, SW480 cells. (B) Forty-eight hours after transfection with a KLF4 expression vector, an immunofluorescence cytochemistry examination for KLF4 and CK20 was conducted. Cells transfected with the KLF4 expression vector expressed both KLF4 and CK20 proteins after nuclear counter-staining with DAPI. Negative control, cells transfected with an empty vector; positive control, HeLa cells. Results shown are representative of three separate experiments. of KLF4 expression induced by bile acids in development of the disease. Other studies have found that KLF4 gene expression has characteristic tissue distribution, with its expression noted in epithelial cells of the gut, skin, and tongue, as well as several other organs.8 9 However, it has not been fully revealed whether KLF4 is expressed in the oesophagus. Herein, we examined the expression level of KLF4 in the oesophagus, and found a lower level as compared to its expression in the small and large intestines, similar to Cdx2. The roles of KLF4 and Cdx2 in the development and carcinogenesis of the intestinal mucosa have been reported to be similar,10 12 thus we examined the role of KLF4 in Barrett’s epithelium development in our study. First, we determined whether bile acids can induce the expression of KLF4 in vivo using Barrett’s epithelium formed in model rats with an oesophago-jejunal anastomosis. KLF4 expression was observed in rat Barrett’s epithelium, thus bile acids are suggested to be inducers of KLF4. However, the expression pattern was found to be different from that of Cdx2, as KLF4-positive cells were observed mainly in the surface villi, whereas there was only a small number of those cells in the crypts. In contrast, Cdx2-positive cells were reported to be abundant in both surface villi and crypts of columnar glands.6 19 20 Although there is a possibility that many different types of cells are present in whole biopsy samples, our results from examinations of human tissues suggest that KLF4 expression is related to the development of Barrett’s oesophagus. Next, we investigated whether bile acids induce KLF4 expression in oesophageal epithelium in vitro. In patients with Barrett’s oesophagus, the most common bile acids found in reﬂuxant are cholic acid, glycocholic acid and taurocholic acid.21 616

Since a mixture of bile acids is considered to provide physiological stimulation,7 21 we investigated the changes in KLF4 promoter activity following stimulation with such a mixture and found that the bile acids stimulated KLF4 promoter activity in three types of oesophageal cells. Furthermore, as expected, KLF4 mRNA and protein expressions in oesophageal epithelial cells were augmented by treatment with the bile acids mixture. Certain bile acids have been reported to be potent activators of NF-kB sites of the promoters of several important proteins, including Cdx1 and Cdx2.6 22 23 The NF-kB family is comprised of several members that interact as homodimers or heterodimers, which function as key regulators of both developmental and pathologic processes. Indeed, deoxycholic acid induced NF-kB subunit p50 nuclear translocation and binding to these sites of the Cdx2 promoter.23 It was also proposed that the Cdx2 promoter is positively regulated by p50 homodimers, whereas it is negatively regulated by p50ep65 heterodimers.24 Furthermore, we and others have suggested that Cdx1 is also an important molecular mediator of Barrett’s metaplasia, and that bile acids stimulate Cdx1 expression by upregulation of p65.22 The present results indicate that the KLF4 promoter is positively regulated by p50 and p65 heterodimers following exposure to bile acids in rat primary cultured keratinocytes. Since activation of the NF-kB pathway is cell-type speciﬁc25 it would be interesting to also evaluate the effects of bile acids on KLF4 expression in cell lines derived from Barrett’s oesophagus. Some homeobox genes, including Cdx1 and Cdx2, have been shown to positively regulate their own expression, as the Cdx promoter has multiple Cdx responsive elements.7 Notably, mechanisms similar to those of Cdx1 and Cdx2 were revealed in regard to KLF4 expression induced by bile acids. KLF4 can be a transcriptional activator or repressor, and it binds to a similar DNA sequence that has either a CACCC homology or is rich in GC contents.14 26 Therefore, we examined whether KLF4 binds to the KLF4 promoter in esophageal epithelial cells. Transfection of the KLF4 expression vector into Het-1A cells increased KLF4 promoter activity and, using a number of deletion constructs of the KLF4 promoter, we conﬁrmed that KLF4 is capable of transactivating the promoter of its own gene through three closely spaced GC boxes within the promoter. Once the expressions of KLF4 are positively regulated by bile acids, even if the induction level is low, the self-replication mechanism induces a higher expression of KLF4.

Figure 9 Schema of mechanism of Barrett’s epithelium development. Bile acids directly stimulate the expressions of KLF4 and Cdx2. Next, auto- and inter-regulation mechanisms between KLF4 and Cdx2 contribute to cellular proliferation and trans-differentiation in intestinal metaplasia. Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Oesophagus In the next step, we investigated the heterologous interregulation mechanism between KLF4 and Cdx2. Prior studies that utilised CHO and RKO cells revealed that Cdx2 induces KLF4 promoter activation via Cdx responsive elements within the KLF4 promoter.14 15 In the present study, transfection of the KLF4 expression vector into Het-1A cells increased the promoter activity of Cdx2 and induced production of Cdx2 protein. Using a number of deletion and mutation constructs, we also revealed that KLF4 protein binds to Sp-1 responsive elements of the Cdx2 promoter. Furthermore, transfection of the Cdx2 expression vector into Het-1A cells increased KLF4 promoter activity and induced KLF4 protein. Taken together, bile acids augment KLF4 expression, and contribute to induce auto- and inter-regulation mechanisms between KLF4 and Cdx2. KLF4 is known to be a direct transcriptional activator of the intestine speciﬁc gene IALP.27 In addition, colonic goblet cells throughout colonic epithelia in KLF4/ mice were found to have reduced expression of MUC2.11 Therefore, we examined whether over-expression of KLF4 in Het-1A cells could trigger their trans-differentiation to intestinal type columnar epithelial cells. Using a number of deletion and mutation constructs, we revealed that KLF4 protein binds to CACCC/Sp-1 responsive elements of the MUC2 promoter. As expected, Het-1A cells transfected with the KLF4 expression vector induced MUC2 mRNA and protein expressions. Furthermore, Het-1A cells transfected with the KLF4 expression vector induced columnar marker CK20. However, KLF4 is a weak inducer of MUC2 as compared to Cdx2, as activation of the MUC2 promoter by a Cdx2 expression vector caused an approximately 10-fold increase, whereas that induced by a KLF4 expression vector was approximately threefold. KLF4 also contributes to induction of an inter-regulation network with Cdx2, and directly and indirectly stimulates cellular trans-differentiation into intestinal metaplasia. Taken together with our previous ﬁndings6 7 these results indicate that over-expression of transcription factors, including KLF4, Cdx2 and Cdx1, induced by bile acids may change the phenotype of oesophageal stem cells into columnar cells (ﬁgure 9). In conclusion, we found that induction of KLF4 expression in oesophageal keratinocytes in response to bile acids has important functions in the induction of metaplastic changes during Barrett’s epithelium development. In addition, our results revealed that the transcriptional network related to KLF4 and Cdx2 has important roles in development of this disease.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

19. 20. 21. 22.

Funding supported in part by grants-in-aid for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan.

23.

Competing interests None.

24.

Ethics approval All rat experimental protocols were approved by the institutional animal care and experimental committee of Shimane University. All human experimental protocols were approved by the ethics committee of Shimane University. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1.

Jankowski JA, Harrison RF, Perry I, et al. Barrett’s esophagus. Lancet 2000;356:2079e85.

Gut 2011;60:608e617. doi:10.1136/gut.2010.221648

25. 26. 27.

Winters C Jr, Spurling TJ, Chobanian SJ, et al. Barrett’s esophagus. A prevalent, occult complication of gastroesophageal reflux disease. Gastroenterology 1987;92:118e24. Gillen P, Keeling P, Byrne PJ, et al. Implication of duodenogastric reflux in the pathogenesis of Barrett’s oesophagus. Br J Surg 1988;75:540e3. Phillips RW, Frierson HF Jr, Moskaluk CA. Cdx2 as a marker of epithelial intestinal differentiation in the esophagus. Am J Surg Pathol 2003;27:1442e7. Moons LMG, Bax DA, Kuipers EJ, et al. The homeodomain protein CDX2 is an early marker of Barrett’s oesophagus. J Clin Patol 2004;57:1063e8. Kazumori H, Ishihara S, Rumi MAK, et al. Bile acids directly augment caudal related homeobox gene Cdx2 expression in oesophageal keratinocytes in Barrett’s epithelium. Gut 2006;55:15e25. Kazumori H, Ishihara S, Kinoshita Y. Roles of caudal-related homeobox gene Cdx1 in oesophageal epithelial cells in Barrett’s epithelium development. Gut 2009;58:620e8. McConnell BB, Ghaleb AM, Nandan MO, et al. The diverse functions of Kru¨ppel-like factors 4 and 5 in epithelial biology and pathobiology. Bioassays 2007;29:549e57. Shields JM, Christy RJ, Yang VW. Identification and characterization of a gene encoding a gut-enriched Kru¨ppel-like factor expressed during growth arrest. J Biol Chem 1996;271:20009e17. Ton-That H, Kaestner KH, Shields JM, et al. Expression of the gut-enriched Kru¨ppel-like factor gene during development and intestinal tumorgenesis. FEBS lett 1997;419:239e43. Katz JP, Perreault N, Goldstein BG, et al. The zinc-finger transcription factor klf4 is required for terminal differentiation of goblet cells in the colon. Development 2002;129:2619e28. Xu J, Lu B, Xu F, et al. Dynamic down-regulation of Kru¨ppel-like factor 4 in colorectal adenoma-carcinoma sequence. J Cancer Res Clin Oncol 2008;134:891e8. Hinoi T, Loda M, Fearon ER. Silencing of cdx2 expression in colon cancer via a dominant repression pathway. J Biol Chem 2003;278:44608e16. Mahatan CS, Kaestner KH, Geiman DE, et al. Characterization of the structure and regulation of the murine gene encoding gut-enriched Kru¨ppel-like factor (Kru¨ppel-like factor 4). Nucleic Acids Res 1999;27:4562e9. Dang DT, Mahatan CS, Dang LH, et al. Expression of the gut-enriched Kru¨ppel-like factor (Kru¨ppel-like factor 4) gene in the human colon cancer cell line RKO is dependent on CDX2. Oncogene 2001;20:4884e90. Levrat M, Lambert R, Kirshbaum G. Esophagitis produced by reflux of duodenal contents in rats. Am J Dig Dis 1962;7:564e73. Aslam F, Palumbo L, Augenlicht LH, et al. The Sp family of transcription factors in the regulation of the human and mouse MUC2 gene promoters. Cancer Res 2001;61:570e6. D’Amico M, Wu K, Fu M, et al. The inhibitor of cyclin-dependent kinase 4a/ alternative reading frame (INK4a/ARF) locus encoded proteins p16INK4a and p19ARF repess cyclin D1 transcription through distinct cis elements. Cancer Res 2004;64:4122e30. Tatsyta M, Mukaisho K, Sugihara H, et al. Expression of Cdx2 in early GRCL of Barrett’s esophagus induced in rats by duodenal reflux. Dig Dis Sci 2005;50:425e31. Pera M, Pera M, de Bolos C, et al. Duodenal-content reflux into the esophasgus leads to expression of Cdx2 and Muc2 in areas of squamous epithelium in rats. Gastrointest Surg 2007;11:869e74. Nehra D, Howell P, Williams CP, et al. Toxic bile acids in gastro-oesophageal reflux disease: influence of gastric acidity. Gut 1999;44:598e602. Wong NA, Wilding J, Bartlett S, et al. CDX1 is an important molecular mediator of Barrett’s metaplasia. Proc Natl Acad Sci USA 2005;102:7565e70. Debruyne PR, Witek M, Gong L, et al. Bile acids induce ectopic expression of intestinal guanylyl cyclase C through nuclear factor-kB and Cdx2 in human esophageal cells. Gastroenterology 2006;130:1191e206. Kim S, Domon-Dell C, Wang Q, et al. PTEN and TNF-a regulation of the intestinal-specific Cdx-2 homeobox gene through a PI3K, PKB/Akt, and NF-kB-dependent pathway. Gastroenterology 2002;123:1163e78. Hormi-Carver K, Zhang X, Zhang HY, et al. Unlike esophageal squamous cells, Barrett’s epithelial cells resist apoptosis by activating the nuclear factor-kB pathway. Cancer Res 2009;69:672e7. Shie J-L, Chen ZY, Fu M, et al. Gut-enriched Kru¨ppel-like factor represses cycline D1 promoter activity through Sp1 motif. Nucleic Acids Res 2000;28:2969e76. Hinnebusch BF, Siddique A, Henderson JW, et al. Enterocyte differentiation marker intestinal alkaline phosphatase is a target gene of the gut-enriched Kru¨ppel-like factor. Am J Physiol 2004;286:G23e30.

617

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Upper GI cancer

Prospective study of serum cysteine levels and oesophageal and gastric cancers in China Gwen Murphy,1 Jin-Hu Fan,2 Steven D Mark,3 Sanford M Dawsey,1 Jacob Selhub,4 Jianbing Wang,2 Philip R Taylor,1 You-Lin Qiao,2 Christian C Abnet1 1

Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA 2 Department of Epidemiology, Cancer Institute, Chinese Academy of Medical Sciences, Beijing, People’s Republic of China 3 Biostatistics & Informatics, University of Colorado, Denver, Colorado, USA 4 Jean Mayer U.S. Department of Agriculture Human Nutrition Research Center on Ageing at Tufts University, Boston, Massachusetts, USA Correspondence to Dr Gwen Murphy, Nutritional Epidemiology Branch, DCEG, National Cancer Institute, 6120 Executive Blvd., EPS 3034, Rockville, MD 20892, USA; [email protected] Prof You-Lin Qiao, Department of Cancer Epidemiology, Cancer Institute, Chinese Academy of Medical Sciences, P.O. Box 2258, Beijing 100021, People’s Republic of China; [email protected] Revised 22 November 2010 Accepted 25 November 2010 Published Online First 17 January 2011

ABSTRACT Background Cancers of the upper gastrointestinal tract remain a significant cause of morbidity and mortality. Cysteine, known to be involved in a myriad of immunomodulatory, anti-oxidant, and anti-carcinogenic pathways, has not been investigated in the aetiology of oesophageal or gastric cancers. To examine the relationship between serum cysteine concentration and risk of these cancers we conducted a nested caseecohort study within the General Population Nutrition Intervention Trial in Linxian, China. Methods 498 oesophageal squamous cell carcinomas (OSCCs) and 255 gastric cardia adenocarcinomas (GCAs) were matched by age and sex to 947 individuals from the wider cohort. We calculated HRs and 95% CIs using the caseecohort estimator for the Cox proportional hazards models, stratified on age and sex, with adjustment for potential confounders. Results Higher concentrations of serum cysteine were significantly associated with a lower risk of both OSCC and GCA. For those in the highest quartile of serum cysteine, compared to those in the lowest, the multivariate HRs were 0.70 for OSCC (95% CI 0.51 to 0.98) and 0.59 for GCA (95% CI 0.38 to 0.91). These associations were dose dependent (p for trend¼0.006 and 0.008, respectively). These inverse associations were not significantly modified by other risk factors, with the exception of age, where a stronger association was noted among persons in the older age strata. Conclusion Higher serum concentrations of cysteine were associated with a significantly reduced risk of OSCC and GCA. Cysteine should be further investigated for its potential as a chemopreventive agent for upper gastrointestinal cancers.

INTRODUCTION Cysteine, a non-essential amino acid and metabolite of methionine, is an essential component of the one-carbon metabolism pathway that has been extensively investigated for its role in the synthesis and methylation of DNA and its possible role in carcinogenesis.1 2 Cysteine can be synthesised by the body under normal physiological conditions, if a sufﬁcient quantity of methionine is available; most high-protein foods, such as pork, chicken, turkey, eggs, milk and yogurt, contain cysteine. Cysteine has many functions besides its role in onecarbon metabolism. A multi-faceted amino acid, cysteine can modulate the immune response by inhibiting activation of nuclear factor-kB (NF-kB) and the expression of several NF-kB-dependent genes.3 It can also stimulate proliferation of lymphocytes and T cell clones under experimental 618

Significance of this study What is already known about this subject?

< Cancers of the stomach and oesophagus remain

a significant cause of morbidity and mortality.

< Cysteine is a non-essential amino acid involved

in immuno-modulatory, anti-oxidant, and anticarcinogenic pathways. < Though cysteine has previously been proposed as a chemopreventive agent, epidemiological studies of it have produced equivocal results.

What are the new findings? < In this large prospective cohort, higher serum

cysteine concentrations are associated with a 30% lower risk of oesophageal squamous cell carcinoma and a 40% lower risk of gastric cardia adenocarcinoma. < The inverse associations between serum cysteine and both oesophageal and gastric cancers are dose dependent.

How might it impact on clinical practice in the foreseeable future? < Cysteine warrants further investigation regarding its potential as a chemopreventive agent for upper gastrointestinal cancers.

conditions.3 Evidence from both human and animal studies suggests that cysteine also exerts direct anti-carcinogenic properties, since it can bind to and detoxify acetaldehyde, a known mutagen and carcinogen.4 5 A recent study involving rats exposed to environmental cigarette smoke showed that orally administered N-acetylcysteine effectively modulated smoke-induced alterations of microRNA expression in rat lungs6; speciﬁcally, N-acetylcysteine modulated microRNAs involved in NF-kBmediated stress and P53 functions. Further, cysteine availability is a rate-limiting step in the formation of glutathione (g-glutamyl-cysteinyl-glycine), an antioxidant tri-peptide with critical roles in nutrient metabolism and wider homeostasis (gene expression, DNA and protein synthesis, cell proliferation and apoptosis, signal transduction, cytokine production and immune response and protein glutathionylation).7 8 Glutathione may also modulate cell death in that high intracellular glutathione levels prevent cell death where low levels enhance it, so that glutathione depletion can activate apoptosis.9 Immune activation may also be modulated by glutathione: it is essential for Gut 2011;60:618e623. doi:10.1136/gut.2010.225854

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Upper GI cancer activation of T lymphocytes, polymorphonuclear leucocytes and cytokine production.10 Cancers of the upper gastrointestinal tract remain a signiﬁcant cause of morbidity and mortality. In 2002 there were an estimated 934 000 new cases of stomach cancer and 462 000 cases of oesophageal cancer, globally, and with a poor prognosis for both cancers, they ranked as the second and sixth leading causes of cancer-related mortality worldwide.11 In Linxian, China, where this study was based, cancer of the upper gastrointestinal tract is endemic; mortality rates for oesophageal cancer in Linxian exceed the Chinese average by 10-fold and the American average (for white men) by 100-fold.12 Cysteine is involved in a range of pathways relevant to carcinogenesis, yet there is a striking dearth of epidemiological studies investigating its role in the aetiology of cancer. To our knowledge, plasma cysteine concentrations have previously been measured only in the context of breast and cervical cancers13e15; given the potential relevance of cysteine to inﬂammatory and other pathways involved in both oesophageal and gastric carcinogenesis, we investigated the association between serum cysteine concentration and risk of oesophageal squamous cell carcinoma and gastric cardia adenocarcinoma in a caseecohort study nested in the General Population Nutrition Intervention Trial in Linxian, China.

METHODS Subjects in this study were selected from the participants of the Linxian General Population Nutrition Intervention Trial. Details of the design, intervention agents, conduct, and primary endpoint analyses of this trial have been published elsewhere.16 17 Brieﬂy, the 29 584 healthy adult trial participants were aged between 40 and 69 years and came from four Linxian communes. One year prior to the intervention each participant was interviewed, given a brief physical examination and had 10 ml of blood drawn. The intervention began in March 1986 and continued through May 1991. Participants were randomly assigned to receive either a vitaminemineral combination or a placebo, in accordance with the partial factorial design of the trial. Four different vitaminemineral combinations were tested: Factor A (10 000 IU of vitamin A and 45 mg of zinc oxide); Factor B (52 mg of riboﬂavin and 40 mg of niacin); Factor C (180 mg of ascorbic acid and 30 mg of molybdenum); and Factor D (50 mg of yeast selenium, 15 mg of b-carotene, and 30 mg of a-tocopherol). Mortality and possible symptoms of cancer among trial participants were ascertained by local village doctors through monthly follow-up, and <1% of the participants were lost to follow-up during the study period. Diagnoses of cancer were made at local commune and county hospitals and supplemented by a study team which provided clinical and diagnostic services, including endoscopy, for patients with symptoms suggestive of oesophageal or gastric cancer. A panel of US and Chinese experts reviewed the diagnostic material for 85% of the cancer cases in this study. The rural setting of the Nutrition Intervention Trial meant that endoscopy was not always possible. As a result, oesophageal cancer cases were frequently proven histologically (45% of cases), although conﬁrmation by barium swallow (23% of cases) or cytology (18%) was also common. Gastric cardia cancer cases were usually proven by histology (68%) of cases, although barium swallow (12%) and cytology (11%) were also used. Cancers in the most proximal 3 cm of the stomach were deﬁned as cardia cancers and those originating elsewhere in the stomach were designated non-cardia cancers. Ninety-ﬁve per cent of the Gut 2011;60:618e623. doi:10.1136/gut.2010.225854

anatomical localisations were made with the use of endoscopy, surgery, and/or x-rays. Senior Chinese diagnosticians conducted reviews for cancer cases without diagnostic materials and for deaths due to causes other than cancer. The median survival times for all cases of oesophageal and gastric cardia cancer during the 5 years of the trial were 0.7 and 0.6 years, respectively. Written informed consent was obtained from each participant prior to enrolment, and human subject protection procedures approved by the institutional review boards of the U.S. National Institutes of Health and the Chinese Academy of Medical Sciences were followed throughout the trial.

Selection of study participants for serum cysteine measurement We used a stratiﬁed caseecohort design,18e20 to select individuals for serum measurements from the cohort of all participants in the General Population Trial. During the 5.25 years of the trial, 640 oesophageal squamous cell carcinomas (OSCCs) and 435 gastric cardia adenocarcinomas (GCAs) were identiﬁed.17 Following a series of simulations and power calculations, 498 OSCCs and 255 GCAs were randomly selected from the identiﬁed cases for inclusion in this analysis (see table 1). Serum cysteine was also measured in a stratiﬁed random sample of all trial participants (n¼947, also selected following power calculations), without regard to outcome, as the comparison group (referred to as the sub-cohort). Six strata were deﬁned by sex and the following age categories (at the start of intervention): <50 years old; 50e60 years old; and >60 years old. Cohort subjects were drawn from each stratum to achieve a greater than 1:1 ratio for control subjects to case subjects for both OSCCs and GCAs combined. The lowest within-stratum ratios of control subjects to case subjects for incident OSCCs or GCAs were 1.4 and 2.7, respectively. Overall, we measured serum cysteine concentrations in 947 subcohort subjects and 753 case subjects.

Laboratory analysis At the time of baseline interview (MarcheMay 1985: 1 year before the initiation of the intervention), a 10 ml venous blood sample was collected from all consenting participants. Samples were stored 3e6 h on ice during transportation to the central ﬁeld station where they were separated, aliquoted and stored at 458C for 3e4 days before being shipped to Beijing for longterm storage at 858C. In 1996, aliquots of these baseline sera were transferred to the National Cancer Institute repository on dry ice. Samples selected for inclusion in the cysteine analysis were thawed, and smaller aliquots were made, immediately refrozen and shipped on dry ice to the analytical laboratories (Jean Mayer U.S. Department of Agriculture Human Nutrition Research Center on Ageing at Tufts University, Boston, Massachusetts, USA). Serum cysteine concentration was determined using a modiﬁed high-performance liquid chromatography method.21 Each batch contained adjacent blinded case and matched control samples, as well as six blinded quality control samples derived from a pool of serum from Linxian residents. The overall coefﬁcient of variation in the 153 QC samples was 13%. Serum creatinine was also measured, and its overall coefﬁcient of variation in the QC samples was 12%.

Statistical analysis Mean and quantile values for serum cysteine were calculated using the sampling weights for the whole General Population Trial cohort, so that estimates are derived from the entire cohort, including those who may have later developed cancer. We used 619

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Upper GI cancer Table 1 Study characteristics of cases and the sub-cohort from the Linxian General Population Nutrition Intervention Trial cohort, overall and by sex Variable

Sub-cohort*

Oesophageal squamous cell carcinoma

Gastric cardia adenocarcinoma

Number, total N (%) Men Women

947 515 (54%) 432 (46%)

498 241 (48%) 257 (53%)

255 167 (65%) 88 (35%)

Age at baseline, mean (SD) Men Women

57 (7.7) 59 (6.8) 56 (8.4)

57 (7.8) 58 (7.0) 55 (8.3)

58 (6.4) 58 (6.1) 56 (6.7)

Cigarette smoking, N (%) yesy Men Women

364 (38%) 362 (70%) 2 (<1%)

190 (38%) 190 (79%) 0 (0%)

118 (46%) 118 (71%) 0 (0%)

Alcohol drinking, N (%) yesz Men Women

200 (21%) 172 (33%) 28 (6%)

112 (23%) 95 (39%) 17 (7%)

58 (23%) 51 (31%) 7 (8%)

BMI (kg/m2), mean (SD) Men Women

21.8 (2.6) 21.5 (2.1) 22.1 (3.1)

21.4 (2.4) 21.6 (2.1) 21.3 (2.6)

21.6 (2.5) 21.3 (2.0) 22.1 (3.2)

Serum creatinine (mg/dl), Mean (SD) Men Women

0.84 (0.18) 0.91 (0.15) 0.75 (0.17)

0.83 (0.20) 0.92 (0.16) 0.75 (0.19)

0.86 (0.22) 0.93 (0.16) 0.72 (0.24)

*Some subjects who later developed one of these cancers are included in the randomly selected sub-cohort. yDaily cigarette consumption among men was generally low and few subjects had quit smoking, so smoking is presented as never smoking versus ever smoking. zAlcohol consumption was minimal at the time of the baseline interview (1985), so alcohol drinking was categorised as any consumption in the previous 12 months versus none

multivariable linear regression to examine the relationships between serum cysteine concentration and its possible predictors. In this linear regression model age was centred at 55 years (the middle of the sub-cohort age range) and body mass index (BMI; weight, in kg/(height, in m)2) was centred at the sub-cohort median, 22 kg/m2. Follow-up time was measured from the start of the intervention, in March 1986, to any cancer diagnosis, loss to follow-up, or death. HRs and 95% CIs were calculated using the caseecohort estimator for the Cox proportional hazards models.18e20 22 All estimates come from models stratiﬁed on the six sex-age sampling strata. Adjustment for variation within age strata was made using stratum-speciﬁc age terms for continuous age. The proportionality assumption was examined using models that allowed time-dependent HRs and no evidence was found that HRs varied with time. The relationship between cysteine and risk of OSCC or GCA was examined using a number of alternative metrics. Continuous estimates of cysteine were scaled to 19 nmol/ml (1/2 the IQR: (75th percentilee25th percentile)O2). HRs were also estimated according to quartiles and ordinal values were assigned to the quartiles for trend tests. Multivariate models were stratiﬁed on the six sex and age group categories used in the subcohort selection. Additional adjustments included separate continuous age variables for each stratum and variables for ever smoking, any alcohol intake in the previous 12 months, BMI (measured height and weight), serum creatinine (mg/dl) and serum creatinine squared (because of a nonlinear association between creatinine and cysteine). Stratiﬁed models were also examined for sex, age, cigarette smoking and alcohol drinking. These stratiﬁed models included an interaction term (cross-product) and were fully adjusted for all other 620

terms. As cigarette and alcohol consumption were almost completely restricted to men in this cohort, these stratiﬁed models are not presented for women. To investigate the possibility the pre-clinical cancers may affect serum cysteine concentrations we examined whether the HR associated with cancer that developed in the ﬁrst 2 years of the study differed from the HR associated with cancer that developed after 2 years. These differences were not signiﬁcant for either OSCC (p¼0.70) or GCA (p¼0.89). The randomised nature of the original intervention meant that confounding by intervention status was unlikely; possible interaction between serum cysteine concentration and the intervention was examined and this was not signiﬁcant.

RESULTS Baseline characteristics of the cases and sub-cohort are shown in table 1. Incident cancer cases included 498 OSCC and 255 GCA. There were almost twice as many GCA cases in men than in women, but the number of OSCC cases was about equal in men and women. The associations between smoking, alcohol and BMI and both gastric and oesophageal cancers have been previously reported for this population.23 The median (and 25th and 75th percentile) serum cysteine concentrations (in nmol/ml) in the cases and cohort participants are shown in table 2. The median serum cysteine concentration for the cohort was 196 nmol/ml (IQR: 178e216 nmol/ml), with no signiﬁcant difference between males (median: 196 nmol/ml; IQR: 179e215 nmol/ml) and females (median: 196 nmol/ml; IQR: 176e217 nmol/ml). With increasing age, the median serum cysteine concentration increased, from 188 nmol/ml (IQR: 168e204 nmol/ml) in the <50 year old category to 203 nmol/ml Gut 2011;60:618e623. doi:10.1136/gut.2010.225854

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Upper GI cancer Table 2 Serum cysteine concentration (nmol/ml) quantiles in the sub-cohort from the Linxian General Population Nutrition Intervention Trial cohort, overall and by sex, age, smoking, and drinking status Quantile General Population Trial Cohort* Overall Sex Men Women Age (in years) <50 50e60 >60 Cigarette smokingy Never smokers Ever smokersy Alcohol drinkingz Non-drinkers Drinkersz

25th

50th

75th

178

196

216

179 176

196 196

215 217

168 186 193

188 203 212

204 225 231

181 179

194 196

219 212

174 183

195 200

216 213

*All concentrations are weighted by the age and sex sampling to reflect the distributions in the full cohort. yDaily cigarette consumption among men was generally low and few subjects had quit smoking, so smoking is presented as never smoking versus ever smoking. zAlcohol consumption was minimal at the time of the baseline interview (1985), so alcohol drinking was categorised as any consumption in the previous 12 months versus none.

(IQR: 186e225 nmol/ml) and 212 nmol/ml (IQR: 193e231 nmol/ ml) in the 50e60 and >60 year old categories. The median cysteine concentrations for OSCC and GCA cases were 191 nmol/ml (IQR: 172e211 nmol/ml) and 198 nmol/ml (IQR: 179e215 nmol/ml), respectively. The paucity of data regarding the predicators of serum cysteine concentrations generally (in any population) led us to investigate such factors in this population (table 3). Age and BMI were signiﬁcantly positively related and serum creatinine level was signiﬁcantly inversely related to serum cysteine concentration. Serum cysteine was signiﬁcantly inversely associated with risk of both OSCC (HR: 0.86; 95% CI: 0.79 to 0.93) and GCA (HR: 0.81; 95% CI: 0.73 to 0.90) after adjusting for age, sex, cigarette smoking, alcohol drinking, BMI, and serum creatinine (table 4). In quartile analysis, risk of both OSCC and GCA decreased signiﬁcantly from the lowest to the highest quartile of serum cysteine (ptrend¼0.006 and ptrend¼0.008, respectively). Those in the top two quartiles of serum cysteine had a signiﬁcantly lower risk of OSCC, relative to those in the lowest quartile (HR: 0.61, 95% CI: 0.44 to 0.86 and HR: 0.70, 95% CI:

Table 3 Predictors of serum cysteine concentration in the sub-cohort from the Linxian General Population Nutrition Intervention Trial cohort, from multivariable linear regression models Variable b coefficient SE p Value Intercept Sex, male Age*, 1 year Cigarette smoking, ever Alcohol drinking, any BMIy, 1 kg/m2 Creatinine, 0.12 units Creatinine squared

242.41 0.020 1.38 4.51 3.54 1.79 13.03 1.019

12.063 2.72 0.1001 2.76 2.30 0.35 3.78 0.29

<0.0001 0.99 <0.0001 0.10 0.13 <0.0001 0.0006 0.0005

*Age was centred at 55 years old in the model, the middle of the sub-cohort age distribution. yBody mass index (BMI) was centred at 22 units, the sub-cohort median.

Gut 2011;60:618e623. doi:10.1136/gut.2010.225854

0.51 to 0.98 for quartiles 3 and 4, respectively). Those in the highest quartile of serum cysteine also had a signiﬁcantly lower risk of GCA, relative to those in the lowest quartile (HR: 0.59, 95% CI: 0.38 to 0.91). Adjusted HRs for risk of OSCC and GCA stratiﬁed by sex, age, cigarette smoking and alcohol drinking are presented in table 5. Signiﬁcant interactions were noted between age and serum cysteine concentration for both OSCC (p¼0.00014) and GCA (p<0.0001); the inverse association between serum cysteine concentration and both tumours was stronger among older people.

DISCUSSION In this large prospective cohort, we found that higher serum cysteine concentrations were associated with signiﬁcantly reduced risk of both OSCC and GCA. Those in the highest quartile of serum cysteine had a 30% lower risk of OSCC and a 40% lower risk of developing GCA relative to those in the lowest quartile. These inverse associations were stronger for those in the older age groups for both cancer sites. Of note, the serum cysteine concentrations reported here (median: 196 nmol/ ml; IQR: 178e196 nmol/ml) were signiﬁcantly lower than those reported for healthy/control subjects in the UK (median: 240 nmol/l; IQR: 216e262 nmol/l)24 and Norway (median: 318 nmol/l; IQR: 289e337 nmol/l).25 This is the ﬁrst study to evaluate the relationship between serum cysteine concentration and risk of oesophageal or gastric cancer. Indeed, we could identify just three studies that have looked at (plasma) cysteine in the context of malignancies at other sites. Most recently, one study examined the association between cysteine, homocysteine, and cancer risk in 812 invasive breast cancer cases and matched controls.14 This group reported a positive association between cysteine and breast cancer risk; those in the highest quintile of plasma cysteine had a RR of 1.65 (95% CI: 1.04 to 2.61) compared to those in the lowest quintile, and the association was dose-dependent (ptrend¼0.04). In contrast, an investigation of 712 incident in situ and invasive breast cancer cases found that women in the highest quintile of plasma cysteine had a signiﬁcantly lower risk of breast cancer relative to those in the lowest quintile (RR: 0.44; 95% CI: 0.26 to 0.74), and this inverse association was dose-dependent (ptrend¼0.002).15 Lastly, a study of 147 women with squamous intraepithelial lesions of the cervix, 185 women with atypical squamous cells of undetermined signiﬁcance, and 191 women with cytologically normal Pap smears reported that women in the highest quartile of plasma cysteine had a signiﬁcantly lower risk of atypical squamous cells of undetermined signiﬁcance than those in the lowest quartile (OR: 0.3; 95% CI: 0.2 to 0.7).13 Given the large number of biological pathways which cysteine is known to modulate, either directly or indirectly, it may not be surprising that it has been both positively and inversely associated with cancer risk in previous studies: it’s role in carcinogenesis is likely determined by both tumour type and physiologic context. While our results suggest some potential for cysteine (or the related N-acteylcysteine) as a chemopreventive agent, evidence from the few reported animal and human studies is mixed. In a rat model where oesophageal papillomas and squamous cell carcinomas and liver tumours were induced, oral administration of N-acetylcysteine signiﬁcantly reduced the multiplicity of oesophageal tumours but had no effect on the incidence or multiplicity of liver tumours.26 A second study, using a rat model of oesophageal adenocarcinoma where tumours were 621

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Upper GI cancer Table 4 HRs and 95% CIs for the association between serum cysteine concentration and risk of oesophageal squamous cell carcinoma and gastric cardia adenocarcinoma in the Linxian General Population Nutrition Intervention Trial cohort Continuousy Group

HR

95% CI

Oesophageal squamous cell carcinoma Age- and sex-adjusted 0.87 0.81 to 0.94 Fully adjusted* 0.86 0.79 to 0.93 Gastric cardia adenocarcinoma Age- and sex-adjusted 0.85 Fully adjusted* 0.81

0.76 to 0.94 0.73 to 0.90

Quartile Q1

Q2

Ref

HR

95% CI

HR

95% CI

HR

95% CI

ptrendz

0.00027 0.00016

1.00 1.00

1.01 0.96

0.77 to 1.34 0.69 to 1.34

0.70 0.61

0.52 to 0.93 0.44 to 0.86

0.73 0.70

0.55 to 0.96 0.51 to 0.98

0.011 0.0064

0.00093 <0.0001

1.00 1.00

1.06 0.95

0.70 to 1.62 0.61 to 1.48

0.96 0.77

0.64 to 1.45 0.50 to 1.19

0.70 0.59

0.46 to 1.06 0.38 to 0.91

0.060 0.0081

p Value

Q3

Q4

*Fully adjusted HRs and 95% CIs in were calculated using models stratified on age and sex, with additional adjustment for continuous age variables within each stratum, cigarette smoking, alcohol drinking, BMI, serum creatinine and serum creatinine squared. yContinuous HRs are scaled to 19 nmol/ml, which is one half of the inter-quartile range in the population distribution (75th percentilee25th percentile O 2) zp for trend comes from a model where quartile of cysteine was entered as an ordinal variable.

induced via oesophagogastroduodenal anastomosis, reported that N-acetylcysteine alone did not signiﬁcantly decrease the incidence of oesophageal adenocarcinomas; however, N-acetylcysteine in combination with a-tocopherol produced a signiﬁcant reduction in tumour incidence.27 Further evidence from animal studies suggests that the dose of N-acetylcysteine may be critical in determining its effect. In a rat mammary cancer model, a low dose of N-acetylcysteine modestly decreased, and a high dose signiﬁcantly increased, tumour occurrence.28 The single study that we have identiﬁed which related to gastric cancer reported that N-acetylcysteine inhibited human gastric cancer SJ-89 cell growth by inducing apoptosis and DNA synthesis arrest.29 A randomised trial of 2592 former or current smokers with head and neck or lung cancers found no effect on survival or the incidence of second primary tumours following 2 years of N-acetylcysteine supplementation.30 On the other hand, a small study of seven volunteers found that carcinogenic acetaldehyde produced by smoking cigarettes could be completely inactivated by sucking a tablet containing 5 mg of 4 L-cysteine, suggesting some potential for cysteine as a chemopreventive agent in certain settings. Without accurate assessment of dietary intake, it is impossible to tell whether serum cysteine may correlate with general

protein intake in the current study population. In another analysis from the same cohort, individuals eating meat more than 12 times a year had a signiﬁcantly reduced risk of OSCC (relative to those who ate meat fewer than four times a year). Higher egg consumption (>36 times/year, relative to two or fewer times/year) was associated with a signiﬁcantly reduced risk of both OSCC and GCA.31 However, it is likely that higher intake of protein (namely meat and eggs) is a proxy for both higher socioeconomic status and higher general nutritional status. Our study has a number of strengths, including the prospective study design, the large number of cancer cases, the availability of data on potential confounders, a state-of-the-art measurement method for serum cysteine concentrations and the virtually complete follow-up of all study participants. Our study also has several limitations. As with any study of this kind, unidentiﬁed, unmeasured confounders could explain the association we report (for table 3 the total model r2¼0.18). The large size of the study, the precision of the laboratory analyses, the close follow-up of trial participants, and the rigorous documentation of cancer diagnoses suggest that neither exposure nor disease misclassiﬁcation are likely to have signiﬁcantly inﬂuenced our estimates. There is always a concern that

Table 5 Adjusted* HRs and 95% CIs for the association between serum cysteine concentration and risk of oesophageal squamous cell carcinoma and gastric cardia adenocarcinoma in the Linxian General Population Nutrition Intervention Trial cohort, stratified by sex, age, cigarette smoking and alcohol drinking Oesophageal squamous cell carcinoma Overall Sex Men Women Age (in years) (<50) (50e60) (>60) Cigarette smokingy Never smokers Ever smokers 0.5 Alcohol drinkingy Non-drinkers Drinkers

HR

95% CI

0.86

0.79 to 0.93

0.89 0.82

0.80 to 0.98 0.72 to 0.92

0.92 0.89 0.80

Gastric cardia adenocarcinoma pinteractiony

pinteractiony

HR

95% CI

0.81

0.73 to 0.89

0.19

0.82 0.77

0.73 to 0.93 0.64 to 0.94

0.66

0.74 to 1.14 0.80 to 1.00 0.70 to 0.91

0.0014

1.10 0.79 0.79

0.76 to 1.58 0.69 to 0.91 0.66 to 0.94

<0.0001

0.91 0.89

0.73 to 1.13 0.79 to 1.00

0.89

0.79 0.84

0.62 to 1.00 0.73 to 0.97

0.41

0.87 0.93

0.76 to 1.00 0.78 to 1.10

0.94

0.77 0.86

0.66 to 0.90 0.69 to 1.07

0.30

*Adjusted HRs and 95% CIs in were calculated using models stratified on age and sex, with additional adjustment for continuous age variables within each stratum, cigarette smoking, alcohol drinking, BMI, serum creatinine and serum creatinine squared. yCigarette and alcohol consumption were almost completely restricted to men, so data stratified by these exposures are not presented for women.

622

Gut 2011;60:618e623. doi:10.1136/gut.2010.225854

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Upper GI cancer pre-clinical disease could lead to alterations in serum measurements, thus creating a misleading association. Biopsy-proven squamous dysplasia is known to be present in about 30% of adults in Linxian.32 Gastric atrophy, as measured by pepsinogen levels, is present, but the prevalence of this ﬁnding depends on the cut-off levels used to deﬁne atrophy (if the deﬁnition of gastric atrophy is a PGI/PGII ratio of <3, <4, <5, or <6, the prevalence of atrophy in this population is 4.5%, 7.5%, 15%, or 26%, respectively.33 To our knowledge, there are no published data on the effect of gastric atrophy/dysplasia or oesophageal dysplasia on cysteine concentrations. In attempting to address this issue in our analysis, we explored the difference in the exposureedisease associations between those with cancers diagnosed within the ﬁrst 2 years of the study and those diagnosed after 2 years and we found no signiﬁcant difference. Cysteine participates in a myriad of biological pathways and the speciﬁc pathways involved in the risk reduction we observed are not clear. The generalisability of our study may be limited due to the speciﬁcities of the population under study (see table 1). In particular, low protein intake in this population34 may be associated with the relatively low range of serum cysteine concentrations which we found. Lastly, it would be interesting to look at the correlation between cysteine and dietary intake of certain foods; however, validated dietary data was not available for this study. In this large prospective population study, we found that individuals with higher baseline serum concentrations of cysteine were associated with a signiﬁcantly reduced risk of both OSCC and GCA. These associations appear biologically plausible. Cysteine should be investigated for its potential as a chemopreventive agent for upper gastrointestinal cancers. Funding This work was supported by the Intramural Research Program, Division of Cancer Epidemiology and Genetics, the National Cancer Institute, at the National Institutes of Health, Department of Health and Human Services.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25. 26.

Competing interests None. Ethics approval This study was conducted with the approval of the US National Institutes of Health and the Chinese Academy of Medical Sciences.

27.

Provenance and peer review Not commissioned; externally peer reviewed.

28.

REFERENCES

29.

1. 2. 3. 4. 5. 6. 7.

Choi S-W, Mason JB. Folate and carcinogenesis: an integrated scheme1-3. J Nutr 2000;130:129e32. Bailey LB, Moyers S, Gregory JF III. Folate. In: Bowman BS, Russell RM, eds. Present Knowledge in Nutrition. Washington, DC: Internation Life Sciences Institute, 2001. Droge W, Eck HP, Gmunder H, et al. Modulation of lymphocyte functions and immune responses by cysteine and cysteine derivatives. Am J Med 1991;91:140Se4S. Salaspuro VJ, Hietala JM, Marvola ML, et al. Eliminating carcinogenic acetaldehyde by cysteine from saliva during smoking. Cancer Epidemiol Biomarkers Prev 2006;15:146e9. Izzotti A, Balansky RM, Dagostini F, et al. Modulation of biomarkers by chemopreventive agents in smoke-exposed rats. Cancer Res 2001;61:2472e9. Izzotti A, Calin GA, Steele VE, et al. Chemoprevention of cigarette smoke-induced alterations of MicroRNA expression in rat lungs. Cancer Prev Res (Phila Pa) 2010;3:62e72. Wu G, Fang YZ, Yang S, et al. Glutathione metabolism and its implications for health. J Nutr 2004;134:489e92.

Gut 2011;60:618e623. doi:10.1136/gut.2010.225854

30.

31. 32. 33. 34.

Hammond CL, Lee TK, Ballatori N. Novel roles for glutathione in gene expression, cell death, and membrane transport of organic solutes. J Hepatol 2001;34:946e54. Celli A, Que FG, Gores GJ, et al. Glutathione depletion is associated with decreased Bcl-2 expression and increased apoptosis in cholangiocytes. Am J Physiol 1998;275:G749e57. Townsend DM, Tew KD, Tapiero H. The importance of glutathione in human disease. Biomed Pharmacother 2003;57:145e55. Parkin DM, Bray F, Ferlay J, et al. Global Cancer Statistics, 2002. CA Cancer J Clin 2005;55:74e108. Blot WJ, Li JY. Some considerations in the design of a nutrition intervention trial in Linxian, People’s Republic of China. Natl Cancer Inst Monogr 1985;69:29e34. Goodman MT, McDuffie K, Hernandez B, et al. Caseecontrol study of plasma folate, homocysteine, vitamin B(12), and cysteine as markers of cervical dysplasia. Cancer 2000;89:376e82. Lin J, Lee I-M, Song Y, et al. Plasma homocysteine and cysteine and risk of breast cancer in women. Cancer Res 2010;70:2397e405. Zhang SM, Willett WC, Selhub J, et al. A prospective study of plasma total cysteine and risk of breast cancer. Cancer Epidemiol Biomarkers Prev 2003;12:1188e93. Li B, Taylor PR, Li JY, et al. Linxian nutrition intervention trials. Design, methods, participant characteristics, and compliance. Ann Epidemiol 1993;3:577e85. Blot WJ, Li JY, Taylor PR, et al. Nutrition intervention trials in Linxian, China: supplementation with specific vitamin/mineral combinations, cancer incidence, and disease-specific mortality in the general population. J Natl Cancer Inst 1993;85:1483e92. Mark SD, Katki H. Influence function based variance estimation and missing data issues in case-cohort studies. Lifetime Data Anal 2001;7:331e44. Prentice RL. A case cohort design for epidemiologic cohort studies and disease prevention trials. Biometrika 1986;73:1e11. Self SG, Prentice RL. Asymptomatic distribution theory and efficiency results for case-cohort studies. Ann Stat 1988;16:64e81. Araki A, Sako Y. Determination of free and total homocysteine in human plasma by high-performance liquid chromatography with fluorescence detection. J Chromatogr 1987;422:43e52. Epicure. Seattle, WA: Hirosoft International Corporation, 1998. Guo W, Blot WJ, Li JY, et al. A nested caseecontrol study of oesophageal and stomach cancers in the Linxian nutrition intervention trial. Int J Epidemiol 1994;23:444e50. Huengsberg M, Waring R, Moffitt D, et al. Serum cysteine levels in HIV infection. AIDS 1998;12:1245. Muller F, Svardal A, Aukrust P, et al. Elevated plasma concentration of reduced homocysteine in patients with human immunodeficiency virus infection. Am J Clin Nutr 1996;63:242e8. Balansky RM, Ganchev G, D’Agostini F, et al. Effects of N-acetylcysteine in an esophageal carcinogenesis model in rats treated with diethylnitrosamine and diethyldithiocarbamate. Int J Cancer 2002;98:493e7. Hao J, Zhang B, Liu B, et al. Effect of alpha-tocopherol, N-acetylcysteine and omeprazole on esophageal adenocarcinoma formation in a rat surgical model. Int J Cancer 2009;124:1270e5. Lubet RA, Steele VE, Eto I, et al. Chemopreventive efficacy of anethole trithione, N-acetyl-L-cysteine, miconazole and phenethylisothiocyanate in the DMBA-induced rat mammary cancer model. Int J Cancer 1997;72:95e101. Li J, Tu HJ, Dai G, et al. N-acetyl cysteine inhibits human signet ring cell gastric cancer cell line (SJ-89) cell growth by inducing apoptosis and DNA synthesis arrest. Eur J Gastroenterol Hepatol 2007;19:769e74. van Zandwijk N, Dalesio O, Pastorino U, et al. EUROSCAN, a randomized trial of vitamin A and N-acetylcysteine in patients with head and neck cancer or lung cancer. For the EUropean Organization for Research and Treatment of Cancer Head and Neck and Lung Cancer Cooperative Groups. J Natl Cancer Inst 2000;92:977e86. Tran GD, Sun XD, Abnet CC, et al. Prospective study of risk factors for esophageal and gastric cancers in the Linxian general population trial cohort in China. Int J Cancer 2005;113:456e63. Pan QJ, Roth MJ, Guo HQ, et al. Cytologic detection of esophageal squamous cell carcinoma and its precursor lesions using balloon samplers and liquid-based cytology in asymptomatic adults in Llinxian, China. Acta Cytol 2008;52:14e23. Ren JS, Kamangar F, Qiao YL, et al. Serum pepsinogens and risk of gastric and oesophageal cancers in the General Population Nutrition Intervention Trial cohort. Gut 2009;58:636e42. Zou XN, Taylor PR, Mark SD, et al. Seasonal variation of food consumption and selected nutrient intake in Linxian, a high risk area for esophageal cancer in China. Int J Vitam Nutr Res 2002;72:375e82.

623

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Helicobacter pylori

Long-term proton pump inhibitor administration worsens atrophic corpus gastritis and promotes adenocarcinoma development in Mongolian gerbils infected with Helicobacter pylori Tadashi Hagiwara, Ken-ichi Mukaisho, Takahisa Nakayama, Hiroyuki Sugihara, Takanori Hattori See Commentary, p 567 Department of Pathology, Shiga University of Medical Science, Shiga, Japan Correspondence to Dr K Mukaisho, Department of Pathology, Shiga University of Medical Science, Seta-tsukinowa-cho, Otsu, Shiga 520-2192, Japan; [email protected]. jp Revised 16 September 2010 Accepted 22 September 2010 Published Online First 20 November 2010

ABSTRACT Background We investigated whether corpus atrophic gastritis worsens in Mongolian gerbils (MGs) after long-term administration of proton pump inhibitor (PPI). MGs are an excellent model for studying Helicobacter pylori-related gastritis and adenocarcinoma. Methods MGs were separated into four groups (n ¼15/ group); H pylori (ATCC43504) was inoculated into the OPZ(omeprazole)+Hp (H pylori) and Hp groups, a PPI (OPZ) was administered to the OPZ+Hp and OPZ groups and the control group received no treatment. MGs had access to food containing omeprazole (100 mg/kg body weight/day) for 6 months, after which their stomachs were removed and cut into nine sections (six sections in the fundus and three sections in the antrum). Corpus atrophy was evaluated by the absence of parietal cells in the six sections in the fundus. First, we calculated a percentage of the area devoid of parietal cells in each haematoxylin and eosin-stained section, and then we scored the degree of atrophy by adding the percentages of the six sections. A full score was 600. Results Neutrophilic and lymphoid infiltrates were greater in the OPZ+Hp group than in the other groups. The corpus atrophy score in the OPZ+Hp group was significantly higher than that in the Hp group (p<0.0048, Student t test). Significantly more adenocarcinomas were found in the OPZ+Hp (60%) than in the Hp (7%) group animals. Conclusion Long-term PPI administration promotes development of adenocarcinoma, which is associated with the progression of atrophic corpus gastritis in MGs infected with H pylori.

INTRODUCTION The use of proton pump inhibitors (PPIs) for treating peptic ulcers and gastro-oesophageal reﬂux disease (GORD) has increased substantially over the past two decades. Furthermore, the widespread use of non-steroidal anti-inﬂammatory drugs in populations at a risk for gastroduodenal side effects is another common indication for PPI therapy.1 2 PPIs are very efﬁcacious but several adverse effects, such as oxyntic cell hyperplasia, glandular cysts, hypergastrinaemia and fundic gland polyps, have been reported.3e5 However, the effect of long-term PPI administration on gastric tumour development remains largely unknown, except for a few reports on carcinoid development in female rats6 and humans.7 It has been widely accepted that gastric 624

Significance of this study What is already known about this subject?

< It has been reported that long-term proton pump

inhibitor (PPI) administration worsens atrophic corpus gastritis in patients infected with Helicobacter pylori. < Infection with H pylori is strongly associated with an increased risk of gastric carcinoma. < Long-term PPI administration leads to hypergastrinaemia, which can ultimately lead to gastric carcinoid formation. < Mongolian gerbils (MGs) are an excellent model for studying H pylori-related gastritis and adenocarcinoma.

What are the new findings?

< Long-term PPI administration promotes devel-

opment of adenocarcinoma in MGs infected with H pylori. < Long-term PPI administration worsens atrophic corpus gastritis in MGs infected with H pylori. < It is suggested that the direct cause of corpus atrophy leading to adenocarcinoma development was the H pylori infection rather than longterm PPI administration. < The risk of corpus atrophy leading to gastric cancer development under long-term PPI administration in patients infected with H pylori might depend on the H pylori strains and/or races.

How might it impact on clinical practice in the foreseeable future? < It is recommended that patients being considered for long-term PPI therapy should be tested for H pylori infection, and if present, this pathogen should be eradicated, especially in Asians infected with the cagA positive-H pylori strain.

adenocarcinoma is associated with chronic atrophic gastritis caused by the presence of Helicobacter pylori.8 9 Also, long-term use of PPIs has been reported to worsen corpus atrophic gastritis in the patients with H pylori infection.10e12 Thus, a potential association exists between gastric neoplasia and the long-term use of PPIs, which has been a topic of debate. Further studies are required Gut 2011;60:624e630. doi:10.1136/gut.2010.207662

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Helicobacter pylori to identify the risk for gastric adenocarcinoma development under long-term PPI administration. However, available cohort studies of chronic PPI users and observational data from large population databases do not provide gastric adenocarcinoma data as an end point.13 Therefore, other methods must be considered to identify the association between PPIs and gastric neoplasia. Animal models are good tools for understanding the pathogenesis of human diseases. However, only a few experimental studies have administered PPIs to laboratory animals for more than 1 month and the method of administration was daily per os or through subcutaneous injection,14 15 but these methods are unsuitable for long-term administration. We recently established an animal model in which long-term administration of PPIs was possible using food containing PPIs and reported the side effects on the long-term use in rats.16 However, rats are not a good species for studying gastritis and adenocarcinoma related to H pylori infection, whereas Mongolian gerbils (MGs) are an excellent model, because the gross and histological ﬁndings mimic lesions induced by H pylori infection in the human gastric mucosa.17 18 The purpose of the present study was to determine whether corpus atrophic gastritis leading to gastric adenocarcinoma in MGs infected with H pylori worsens after long-term administration of PPIs.

MATERIALS AND METHODS All procedures complied with the ethical guidelines for animal experimentation and the care and use of laboratory animals at Shiga University of Medical Science, Japan.

Animals Totally sixty male MGs were used in this experiments. Thirty MGs were infected with H pylori (ATCC43504: both cagA and vacA were positive) (MGS/Sea; Kudo, Inc., Kumamoto, Japan), and 30 were used without H pylori infection. The inoculation dose was 3.23108 CFU/ml and 0.5 ml/head. The animals were housed in an air-conditioned biohazard room designed for infectious animals with a 12-h light/12-h dark cycle. They were provided rodent diet (CE-2, CLEA, Osaka) and water ad libitum.

PPI administration The MGs were separated into four groups (n ¼15/group). One month after birth, MGs from groups OPZ(omeprazole)+Hp and Hp were inoculated with H pylori. OPZ was administered to groups OPZ+Hp and OPZ, and the control group received no treatment. MGs had access to food containing OPZ (100 mg/kg body weight/day) for 6 months, according to a recently established protocol.16 Pellet diet containing 0.1125% OPZ was provided 6 months. OPZ was administered beginning 6 months after infection, when pangastritis started to develop in the H pylori-infected MGs (ﬁgure 1). MGs were sacriﬁced using an overdose of diethyl ether after the OPZ treatment. They were deprived of food but allowed free access to tap water 18 h before sacriﬁce. Immediately after sacriﬁce, blood was drawn from the heart for analyses of serum gastrin levels, and the stomach was quickly removed.

Histological examination The stomach was ﬁxed with 10% formalin in phosphate-buffered saline for 4 h and was cut into nine sections (six sections in the fundus and three sections in the antrum). The tissues were embedded in parafﬁn and cut into 4 mm sections. The sections Gut 2011;60:624e630. doi:10.1136/gut.2010.207662

OPZ+Hp Group (PPI+ Hp+)

H p G r ou p

(PPI- Hp+)

OPZ Group

(PPI+ Hp-)

Control Group (PPI- Hp-)

0 1

7 month

13

Figure 1 Experimental design. Mongolian gerbils were separated into four groups (n ¼15/group), and Helicobacter pylori was inoculated into the OPZ+Hp and Hp groups 1-month after birth (red arrow). Omeprazole (OPZ) was administered to the OPZ+Hp and OPZ groups, starting 6 months after infection (yellow arrow). The control group received no treatment. were stained with haematoxylin&eosin (H&E) to examine stomach morphology.

Evaluation of atrophy A few pyloric-type glands were present in the corpus of the normal MG stomachs immediately distal to the squamous portion. These glands might have been miscounted as corpus atrophic glands; however, there were only a few glands in normal MG stomachs, so the possibility that corpus atrophy scoring was altered was minimal. We ignored the pyloric-type glands in the corpus immediately distal to the squamous portion and evaluated corpus atrophy by the total absence of parietal cells in the six sections in the fundus. This deﬁnition was simple but very strict. We ﬁrst calculated a percentage of the area devoid of parietal cells in each H& E-stained section, and then calculated the sum of the percentages of the six sections. A score of 600 was the full score in one animal (ﬁgure 2).

Definition of adenocarcinoma distinguished from the heterotopic gland and intestinal metaplasia that developed in the submucosa Franco et al reported dysplastic foci and adenocarcinoma in a model using MGs infected with the in vivo-adapted H pylori strain 7.13.19 They diagnosed dysplasia and adenocarcinoma using morphological criteria established previously for gastrointestinal neoplasia in diseased mouse models.20 We therefore also diagnosed adenocarcinoma according to these studies.19 20 The atypical glands, called dysplastic foci in the study by Franco (evident in Franco’s photomicrograph), were detected frequently within the mucosa of our model in the Hp and OPZ+Hp groups. However, the lesion was difﬁcult to distinguish from regenerative atypia due to severe active inﬂammation. We therefore counted only the heterotopic gland lesions and obvious adenocarcinoma in the present study. We deﬁned submucosal dilated glands without nuclear atypia as heterotopic glands. Adenocarcinoma was characterised by marked cellular pleomorphism, cellular atypia and euchromatic nuclei that exhibited bizarre morphological features.20 In a study using MGs infected with H pylori, the histological type of adenocarcinoma was mucinous adenocarcinoma17 and well-differentiated tubular adenocarcinoma had also developed.20e22 The adenocarcinoma must accompany invasive growth and not glandular herniation. The criteria to discriminate between glandular herniation and true invasion were also determined as described previously.19 625

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Helicobacter pylori Figure 2 Corpus atrophy scores. Each bold line shows the area of atrophy. Corpus atrophy was evaluated by the absence of parietal cells. We calculated the percentage of the area devoid of parietal cells in each haematoxylin and eosin-stained section (A). The figure shows samples of (i) 20%, (ii) 80% and (iii) 100% atrophy in the fundic gland section of an animal inoculated with Helicobacter pylori and administered omeprazole 6 months later (OPZ+Hp) (B).

Immunohistochemical analyses for BrdU

59 -Bromo-29 -deoxyuridine (BrdU) at a dose of 100 mg/kg was injected intraperitoneally into all MGs at 60 min before sacriﬁce. Immunohistochemical staining was conducted using BrdU. For immunohistochemistry, the streptavidinebiotin method was performed using a Histoﬁne kit (Nichirei, Tokyo, Japan). Peroxidase binding sites were visualised by the diaminobenzidine method, and the sections were lightly counterstained with haematoxylin. The numbers of BrdU-labelled cells in two glands of adenocarcinoma and the heterotopic gland in the Hp and OPZ+Hp groups were counted microscopically, and the mean percentages of positive cells among the total cells in the two glands of adenocarcinoma and heterotopic gland were determined.

Serum gastrin levels The serum gastrin levels were measured by radioimmunoassay (Medic, Yasu, Japan).

Statistical analysis The adenocarcinoma and heterotopic gland indices were compared by Welch’s t test. The corpus atrophy score was expressed as mean6SE, and the groups were compared by the Student t test and Fisher’s exact test. p<0.05 was considered statistically signiﬁcant.

RESULTS Body weight The body weights in the OPZ+Hp, Hp, OPZ and control groups were 101.3610.6, 97.0611.5, 104.365.3 and 120.069.9 g, respectively (table 1). No signiﬁcant differences in weight were observed among the groups.

Serum gastrin level The gastrin level in the OPZ+Hp, Hp, OPZ and control groups were 349.36181.3, 213.06122.1, 80.1613.5 and 97.3624.6 pg/ ml, respectively (table 1). The gastrin level in the OPZ+Hp group was signiﬁcantly higher than that in the OPZ and control (p<0.008) groups, and the gastrin level in the OPZ+Hp group tended to be higher than that in the Hp group (p<0.0877, Student t test; table 1).

Macroscopic findings No macroscopic abnormalities were found in the control group. The stomach wall was mildly thickened in the OPZ group 626

compared to that in the control group. Ruggedness of the superﬁcial mucosa in the border lesion between the corpus and the antrum was detected in the Hp group. Polypoid lesions and a marked reduction in the normal corpus area were observed in the OPZ+Hp group (ﬁgure 3).

Microscopic findings Inflammatory cell infiltrates We evaluated the histological features in the centre of the fundus. The mucosa was thicker in the OPZ than in the control group, and inﬁltration of neutrophils and lymphocytes into the mucosa was detected in the OPZ+Hp and Hp groups. Corpus atrophy included replacement of the oxyntic cells by mucoustype cells in the OPZ+Hp and Hp groups. Marked foveolar hyperplasia was also noted in these atrophic glands. Macroscopical extensive polypoid formations were caused by foveolar hyperplastic changes, cystic dilated glands and severe inﬁltration of neutrophils and lymphocytes. The degree of glandular atrophy, deﬁned as the total loss of parietal cells and of neutrophilic and lymphocytic inﬁltration, was much higher in the OPZ+Hp group than in the control group (ﬁgure 4).

Incidence of intestinal metaplasia, heterotopic glands and adenocarcinoma Based on the analyses of the sections, we counted the animals with intestinal metaplasia, heterotopic glands and adenocarcinoma (table 2). Intestinal metaplasia was detected in the OPZ +Hp (14 of 15) and Hp (13 of 15) groups, but not in the OPZ or control groups (table 2). Heterotopic glands were identiﬁed in the submucosal layer from the OPZ+Hp (15 of 15), Hp (14 of 15) and OPZ (1 of 15) groups but not in the control group (ﬁgure 5AeC Table 1 Body weight, serum gastrin level and corpus atrophy score. No significant differences in body weight were noted among the groups. The serum gastrin level in the OPZ+Hp group was significantly higher than that in the OPZ and control groups (p <0.008). The gastrin level in the OPZ+Hp group was slightly higher than that of the Hp group (p <0.0877). The corpus atrophy score in the OPZ+Hp group was significantly higher than that in the Hp group (p <0.0008, Student t test)

OPZ+Hp group Hp group OPZ group Control group

Body weight (g)

Serum gastrin level (pg/ml)

Corpus atrophy score (/600)

101.3610.6 97.0611.5 104.365.3 120.069.9

349.36181.3 213.06122.1 80.1613.5 97.3624.6

292.8623.0* 151.2630.1* 0.060.0 0.060.0

*p<0.0008, Student t test.

Gut 2011;60:624e630. doi:10.1136/gut.2010.207662

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Helicobacter pylori

Figure 3 Macroscopic findings. Polypoid lesions and a marked reduction in the normal corpus area were observed in the OPZ+Hp group (A). The polypoid lesions were milder in the Hp group than in OPZ +Hp group (B). A mild thickening of the stomach wall was observed in the OPZ group (C) compared to that in the control group (D). The OPZ +Hp and Hp groups were inoculated with Helicobacter pylori (Hp), and the OPZ and OPZ+Hp groups were administered the proton pump inhibitor omeprazole (OPZ) 6 months later. The control group received no treatment. table 2). Adenocarcinomas were detected in the OPZ+Hp (9 of 15) and Hp (1 of 15) groups, but not in the OPZ or control groups. The adenocarcinomas detected were exclusively of the mucinous type; detached mucous columnar epithelia ﬂoating in mucin lakes showed structural and nuclear atypism and several mitotic cells (ﬁgure 5DeI). The incidence of adenocarcinomas was greater in the OPZ+Hp group than in the Hp group (p<0.0052, Fisher’s exact test) (table 2).

BrdU labelling index (%) The BrdU labelling index of the adenocarcinoma (21.262.06) was signiﬁcantly higher than that of the heterotopic gland (1.2160.26) (p <0.0001) (ﬁgure 6).

Corpus atrophy scores More severe neutrophilic and lymphoid inﬁltrates were detected microscopically in the OPZ+Hp group than in the other groups (ﬁgure 4). No glandular atrophy was noted in the OPZ and control groups. The corpus atrophy score in the OPZ+Hp group was signiﬁcantly higher than that in the Hp group (p<0.0008, Student t test) (table 1).

DISCUSSION Kuipers et al11 hypothesised that inﬂammation of the corpus worsens because stomach pH rises in patients infected with H pylori who are undergoing long-term PPI therapy. This may cause an extension of H pylori infection from the pylorus to the corpus. Furthermore, interleukin (IL) 1b and tumour necrosis factor (TNF) a are proinﬂammatory cytokines with potent acidsuppressive properties, and highly expressed polymorphisms within the human IL-1b and TNFa gene promoters heighten the risk for gastric adenocarcinoma among subjects infected with Gut 2011;60:624e630. doi:10.1136/gut.2010.207662

Figure 4 Inflammatory cell infiltrate. We evaluated the histological features in the same area of the fundus centre. Group OPZ+Hp animals had corpus atrophy and marked neutrophilic and lymphocytic infiltrates (A). Small numbers of neutrophils and lymphocyte infiltrates were observed in the mucosa of the Hp group (B). Mild hyperplasia of the superficial mucosa was observed in the OPZ group (C). No pathological findings were observed in the control group (D). H pylori.23 24 However, a mechanism for this change remains unclear. In the present study, we conﬁrmed Kuipers’ hypothesis using a well-characterised MG model of H pylori-associated gastritis and gastric adenocarcinoma; long-term use of PPIs worsens corpus atrophic gastritis in patients with H pylori infection.11 Furthermore, the results also suggest that long-term PPI administration promotes adenocarcinoma development in the corpus. However, this study has two main limitations. One concerns the strains of MGs and H pylori, and the other concerns the PPI dose. Long-term H pylori infection in MGs can eventually progress to gastric adenocarcinoma25 26; therefore, this prolonged time course precludes large-scale analyses to evaluate the effects of both pathogen and host in a carcinogenic cascade. H pyloriinduced cancer in this model has never been reported outside Japan or China.22 Watanabe et al17 used the TN2GF4 strain and the other groups used the ATCC43504 strain. These strains are positive for both cagA and vacA. One strain-speciﬁc H pylori constituent that augments cancer risk is the cag pathogenicity island,9 a genetic locus that encodes a type IV secretion system. Upon delivery into host cells by the cag secretion system, the terminal gene product in the island, CagA, undergoes Src-dependent tyrosine phosphorylation and activates Table 2 Cases of intestinal metaplasia, heterotopic glands and adenocarcinoma. Significantly more adenocarcinomas were detected in the OPZ+Hp group than in the Hp group

OPZ+Hp group Hp group OPZ group Control group

Corpus gastritis

Intestinal metaplasia

Heterotopic glands

Adenocarcinoma

15/15

14/15

15/15

9/15*

15/15 0/15 0/15

13/15 0/15 0/15

14/15 1/15 0/15

1/15* 0/15 0/15

*p<0.0052 (Fisher’s exact test).

627

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Helicobacter pylori Figure 5 Comparison between heterotopic glands and adenocarcinomas. (AeC) Heterotopic glands, (DeI) adenocarcinomas, (B, C) Higher magnification of figure 5A, (E, F) Higher magnification of figure 5D, (I) Higher magnification of figure 5H. Heterotopic glands were found in the submucosal area in the OPZ+Hp, Hp and OPZ groups. These glands had no nuclear pleomorphism or structural atypia (A, B, C). The gland was covered with a muscular mucosae component, and a desmoplastic response was absent (B, C). In contrast, adenocarcinomas were detected in the OPZ+Hp (9 of 15) and Hp (1 of 15) groups. The adenocarcinoma revealed nuclear pleomorphism and several mitotic cells (DeI). Obvious desmoplastic reaction was observed around the atypical glands (D, E, H, I). Many atypical glands were detected in the submucosal layer (DeI).

a eukaryotic phosphatase (SHP-2), leading to dephosphorylation of host cell proteins and cellular morphological changes.27e29 Translocation, but not phosphorylation, of CagA also disrupts the apicalejunctional complexes, resulting in a loss of cellular polarity.30 Therefore, the cagA-positive strain is likely critical for gastric cancer development. However, in the previous study, adenocarcinoma did not develop among any of the infected MGs using the cag+vacA s1a human H pylori gastric ulcer strain B128.19 Thus, the cause of cancer development may depend on cagA as well as on host factors that inﬂuence gastric carcinogenesis. Moreover, we also considered the MG strains. In the literature, it has been reported that the MG strain in Europe and the US were separated from those in Japan in 1954.31 The MG strains in Asian countries may be different from those in Europe and the US. From these ﬁndings, even if a patient is infected with H pylori, the risk of corpus atrophy leading to gastric cancer development under long-term PPI administration might depend on the H pylori strains and/or races. The other potential problem concerns the OPZ dose. In the literature, the method of OPZ administration is daily per os or

subcutaneous injection until we established a method for longterm PPI administration using OPZ-containing food.14 15 The OPZ dose in the present study was 100 times more than the usual therapeutic dose used in humans. However, it is unclear how much of the dose is inactivated during the process of incorporating OPZ in food. A dose of 100 mg/kg/day was determined from the results of a previous preliminary study using a rat model.16 The MGs were healthy in the OPZ group, and neither glandular atrophy nor tumour development was detected in their stomachs. These ﬁndings suggest that the dose was appropriate for long-term PPI administration and that the cause of gastric atrophy leading to cancer development is not a large OPZ dose. We understand the difﬁculty of diagnosing adenocarcinoma in animal models. Elfvin et al25 stated that adenocarcinoma was not found in any MGs when they examined sequential morphological changes in the stomach for up to 62 weeks using MGs infected with TN2GF4 and SS1. They suggested that the glands were buried in the submucosal layer, that these changes might be misinterpreted as adenocarcinoma and that glands in the muscularis propria were insufﬁcient to diagnose gastric

Figure 6 Immunohistochemical stainings of BrdU in adenocarcinomatous and heterotopic glands. (A) Heterotopic glands, (B) adenocarcinomas. Each photo of these glands was taken from the OPZ+Hp group. The BrdU adenocarcinoma labelling indices (21.262.06) (A) were significantly higher than those of the heterotopic gland (1.2160.26) (B) (p <0.0001).

628

Gut 2011;60:624e630. doi:10.1136/gut.2010.207662

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Helicobacter pylori adenocarcinoma, because the deranged glandular structures can grow in and below the submucosa. Therefore, we diagnosed adenocarcinoma and distinguished it from heterotopic glands. Adenocarcinomas found in the present study were mostly of the mucinous type; mucin lakes contained detached mucous columnar epithelia with nuclear and structural atypia and several mitotic cells. Such a mucinous lesion, quite different from those of the heterotopic glands and intestinal metaplasia developed in submucosa, were seldom detected in previous studies after a 1-year infection. We also found that the BrdU labelling indices of mucinous and tubular adenocarcinoma were signiﬁcantly higher than those of the heterotopic glands. Adenocarcinomatous glands have a much higher proliferative capacity. These ﬁndings suggest that adenocarcinoma diagnosed in the present study was completely different from both heterotopic glands and intestinal metaplasia developed in the submucosa. Profound acid suppression therapy leads to hypergastrinaemia in nearly all patients receiving long-term administration of PPIs.26 Furthermore, prolonged hypergastrinaemia leads to hyperplasia of enterochromafﬁn-like cells, which can ultimately lead to gastric carcinoid formation.32 In humans, diffuse, linear or micronodular hyperplasia of enterochromafﬁn-like cells is observed in 10e30% of chronic PPI users, particularly in H pyloriinfected patients with markedly increased gastrin levels.33 Dysplasia or invasive carcinoid formation has not been described in long-term PPI users; thus, it is not an indication for surveillance in PPI maintenance users. Insulinegastrin transgenic mice, which show hypergastrinaemia with H felis or H pylori SS1 infection, progress to gastric cancer.34 35 In the present study, the serum gastrin levels in the OPZ+Hp group, which was administered OPZ and infected with H pylori, were higher than those of the other groups, and adenocarcinomas were signiﬁcantly more common in the OPZ+Hp group than in the other groups. These ﬁndings suggest that hypergastrinaemia might promote the development of gastric cancer. One of the most important ﬁndings in the present study was that long-term OPZ administration without H pylori infection did not induce corpus atrophy. No neoplastic lesions developed in a previous study using male rats administered OPZ for 1 year.36 These ﬁndings suggest that the direct cause of corpus atrophy leading to the development of adenocarcinoma was the H pylori infection rather than long-term PPI administration. H pylori eradication induces regression of gastric mucosal inﬂammation and atrophy in chronic PPI users without affecting the efﬁcacy of PPI therapy for GORD.37 Therefore, the advice of the Maastricht consensus panel was to eradicate H pylori in infected subjects requiring long-term maintenance treatment with a PPI.38 H pylori eradication makes long-term PPI treatment a relatively safe therapy for patients with peptic acid disorders and GORD. Taken together, although the present study had several limitations, we were able to show that long-term PPI administration worsens atrophic corpus gastritis and promotes the development of adenocarcinoma in MGs infected with H pylori. The main cause of corpus atrophy was not the PPI but H pylori infection. H pylori might move to the oral side under decreased gastric acid conditions, induce corpus atrophic gastritis and promote adenocarcinoma development.13 Hypergastrinaemia may also be associated with gastric cancer development. Thus, it is recommended that patients being considered for long-term PPI therapy should be tested for H pylori infection, and if present, this pathogen should be eradicated, especially in Asians infected with the cagA positive H pylori strain. Gut 2011;60:624e630. doi:10.1136/gut.2010.207662

Competing interests None. Ethics approval All procedures complied with the ethical guidelines for animal experimentation and the care and use of laboratory animals at Shiga University of Medical Science, Japan. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2.

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

Targownik LE, Metge CJ, Leung S, et al. The relative efficacies of gastroprotective strategies in chronic users of nonsteroidal anti-inflammatory drugs. Gastroenterology 2008;134:937e44. Hawkey CJ, Karrasch JA, Szczepan˜ski L, et al. Omeprazole compared with misoprostol for ulcers associated with nonsteroidal antiinflammatory drugs. Omeprazole versus Misoprostol for NSAID-induced Ulcer Management (OMNIUM) Study Group. N Engl J Med 1998;338:727e34. Jalving M, Koornstra JJ, Wesseling J, et al. Increased risk of fundic gland polyps during long-term proton pump inhibitor therapy. Aliment Pharmacol Ther 2006;24:1341e8. Stolte M. Fundic gland polyps: a rare, innocuous, and reversible disturbance. Gastroenterology 1993;105:1590e1. Raghunath AS, O’Morain C, McLoughlin RC. Review article: the long-term use of proton-pump inhibitors. Aliment Pharmacol Ther 2005;22:55e63. Havu N. Enterochromaffin-like cell carcinoids of gastric mucosa in rats after life-long inhibition of gastric secretion. Digestion 1986;35:42e55. Dawson R, Manson JM. Omeprazole in oesophageal reflux disease. Lancet 2000;356:1770e1. Correa P. Chronic gastritis as a cancer precursor. Scand J Gastroenterol Suppl 1984;104:131e6. Peek RM Jr, Blaser MJ. Helicobacter pylori and gastrointestinal tract adenocarcinomas. Nat Rev Cancer 2002;2:28e37. Kuipers EJ, Uyterlinde AM, Pen˜a AS, et al. Increase of Helicobacter pyloriassociated corpus gastritis during acid suppressive therapy: implications for longterm safety. Am J Gastroenterol 1995;90:1401e6. Kuipers EJ, Lundell L, Klinkenberg-Knol EC, et al. Atrophic gastritis and Helicobacter pylori infection in patients with reflux esophagitis treated with omeprazole or fundoplication. N Engl J Med 1996;334:1018e22. Garcı´a Rodrı´guez LA, Lagergren J, Lindblad M. Gastric acid suppression and risk of oesophageal and gastric adenocarcinoma: a nested case control study in the UK. Gut 2006;55:1538e44. Kuipers EJ. Proton pump inhibitors and gastric neoplasia. Gut 2006;55;1217e21. Viste A, Øvrebø K, Maartmann-Moe H, et al. Lanzoprazole promotes gastric carcinogenesis in rats with duodenogastric reflux. Gastric Cancer 2004;7:31e5. Sundler F, Andersson K, Mattsson H, et al. Administration of omeprazole to rats for one year produces reciprocal effects on antral gastrin and somatostatin cells and no effect on endocrine cells in the colon. Digestion 1995;56:194e8. Hagiwara T, Mukaisho K, Ling ZQ, et al. Development of pancreatic acinar cell metaplasia after successful administration of omeprazole for 6 months in rats. Dig Dis Sci 2007;52:1219e24. Watanabe T, Tada M, Nagai H, et al. Helicobacter pylori infection induces gastric cancer in mongolian gerbils. Gastroenterology 1998;115:642e8. Kudo T, Lu H, Wu JY, et al. Pattern of transcription factor activation in Helicobacter pylori-infected Mongolian gerbils. Gastroenterology 2007;132:1024e38. Franco AT, Israel DA, Washington MK, et al. Activation of beta-catenin by carcinogenic Helicobacter pylori. Proc Natl Acad Sci USA 2005;102:10646e51. Boivin GP, Washington K, Yang K, et al. Pathology of mouse models of intestinal cancer: consensus report and recommendations. Gastroenterology 2003;124:762e77. Honda S, Fujioka T, Tokieda T, et al. Development of Helicobacter pylori-induced gastric carcinoma in Mongolian gerbils. Cancer Research 1998;58:4255e9. Zheng Q, Chen XY, Shi Y, et al. Development of gastric adenocarcinoma in Mongolian gerbils after long-term infection with Helicobacter pylori. J Gastroenterol Hepatol 2004;19:1192e8. El-Omar EM, Carrington M, Chow WH, et al. Interleukin-1 polymorphisms associated with increased risk of gastric cancer. Nature 2000;404:398e402. El-Omar EM, Rabkin CS, Gammon MD, et al. Increased risk of noncardia gastric cancer associated with proinflammatory cytokine gene polymorphisms. Gastroenterology 2003;124:1193e201. Elfvin A, Bo¨lin I, Von Bothmer C, et al. Helicobacter pylori induces gastritis and intestinal metaplasia but no gastric adenocarcinoma in Mongolian gerbils. Scand J Gastroenterol 2005;40:1313e20. Orlando LA, Lenard L, Orlando RC. Chronic hypergastrinemia: causes and consequences. Dig Dis Sci 2007;52:2482e9. Odenbreit S, Pu¨ls J, Sedlmaier B, et al. Translocation of Helicobacter pylori CagA into gastric epithelial cells by type IV secretion. Science 2000;287:1497e500. Selbach M, Moese S, Hauck CR, et al. Src is the kinase of the Helicobacter pylori CagA protein in vitro and in vivo. J Biol Chem 2002;277:6775e8. Higashi H, Tsutsumi R, Muto S, et al. SHP-2 tyrosine phosphatase as an intracellular target of Helicobacter pylori CagA protein. Science 2002;295:683e6. Amieva MR, Vogelmann R, Covacci A, et al. Disruption of the epithelial apicalejunctional complex by Helicobacter pylori CagA. Science 2003;300:1430e4. Schwentker V. The gerbil, a new laboratory animal. The Illinos Vet 1963;6:5e9. Larsson H, Carlsson E, Hakanson R, et al. Time-course of development and reversal of gastric endocrine cell hyperplasia after inhibition of acid secretion. Studies with

629

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Helicobacter pylori

33. 34. 35.

omeprazole and ranitidine in intact and antrectomized rats. Gastroenterology 1988;95:1477e86. Klinkenberg-Knol EC, Nelis F, Dent J, et al. Long-term omeprazole treatment in resistant gastroesophageal reflux disease: efficacy, safety and influence on gastric mucosa. Gastroenterology 2000;118:661e9. Wang TC, Dangler CA, Chen D, et al. Synergistic interaction between hypergastrinemia and Helicobacter infection in a mouse model of gastric cancer. Gastroenterology 2000;118:36e47. Fox JG, Rogers AB, Ihrig M, et al. Helicobacter pylori-associated gastric cancer in INS-GAS mice is gender specific. Cancer Res 2003;63:942e50.

36. 37.

38.

Hagiwara T, Mukaisho K, Ling ZQ, et al. Rebamipide contributes to reducing adverse effects of long-term administration of omeprazole in rats. Dig Dis Sci 2007;52:988e94. Kuipers EJ, Nelis GF, Klinkenberg-Knol EC, et al. Cure of Helicobacter pylori infection in patients with reflux oesophagitis treated with long-term omeprazole reverses gastritis without exacerbation of reflux disease; results of a randomised controlled trial. Gut 2004;53:12e20. Malfertheiner P, Megraud F, O’Morain C, et al. Current concepts in the management of Helicobacter pylori infection: the Maastricht III Consensus Report. Gut 2007;56:772e81.

Editor’s quiz: GI snapshot A prematurely terminated journey CLINICAL PRESENTATION A 50-year-old man presented to our emergency department because of worsening abdominal pain of 2 days duration accompanied by fever. Upon admission the middle and lower abdomen was moderately painful upon palpation. His temperature was 38.88C. Laboratory tests showed leucocytosis (15 3 109/l) and thrombocytopenia (61 3 109/l), a C-reactive protein level of 276 mg/l (normal <5), and slightly elevated liver enzymes. An abdominal CT scan with intravenous contrast was performed. (ﬁgures 1e4).

Figure 3

Abdominal CT scan no.3.

Figure 4

Abdominal CT scan, section through the liver.

QUESTION What does the CT scan show? See page 687 for the answer

Figure 1 Abdominal CT scan no.1.

S A Mu¨ller-Lissner,1 E Fuhrmann2 1

Department of Internal Medicine, Park-Klinik Weissensee, Berlin, Germany; Department of Radiology, Park-Klinik Weissensee, Berlin, Germany

2

Correspondence to Professor Stefan A. Mu¨ller-Lissner, Department of Internal Medicine, Park-Klinik Weissensee, Schoenstrasse 80, D-13086 Berlin, Germany; [email protected] Competing interests None. Patient consent Obtained. Provenance and peer review Not commissioned; externally peer reviewed. Published Online First 14 May 2010

Figure 2 Abdominal CT scan no.2.

630

Gut 2011;60:630. doi:10.1136/gut.2009.190546

Gut May 2011 Vol 60 No 5

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Inflammatory bowel disease

Dysbiosis of the faecal microbiota in patients with Crohn’s disease and their unaffected relatives Marie Joossens,1 Geert Huys,2,3 Margo Cnockaert,2 Vicky De Preter,1 Kristin Verbeke,1 Paul Rutgeerts,1 Peter Vandamme,2 Severine Vermeire1 < Additional tables are

published online only. To view these files please visit the journal online (http://gut.bmj. com). 1

Department of Gastroenterology, Gastroenterology and Leuven Food Science and Nutrition Centre (LFoRCe), University Hospital Gasthuisberg, Leuven, Belgium 2 Laboratory of Microbiology, Ghent University, Ghent, Belgium 3 BCCM/LMG Bacteria Collection, Faculty of Sciences, Ghent University, Ghent, Belgium Correspondence to Se´verine Vermeire, Department of Gastroenterology, University Hospital Gasthuisberg, Herestraat 49, 3000 Leuven, Belgium; severine.vermeire@ uz.kuleuven.ac.be Accepted 5 November 2010 Published Online First 5 January 2011

ABSTRACT Background and aims A general dysbiosis of the intestinal microbiota has been established in patients with Crohn’s disease (CD), but a systematic characterisation of this dysbiosis is lacking. Therefore the composition of the predominant faecal microbiota of patients with CD was studied in comparison with the predominant composition in unaffected controls. Whether dysbiosis is present in relatives of patients CD was also examined. Methods Focusing on families with at least three members affected with CD, faecal samples of 68 patients with CD, 84 of their unaffected relatives and 55 matched controls were subjected to community fingerprinting of the predominant microbiota using denaturing gradient gel electrophoresis (DGGE). To analyse the DGGE profiles, BioNumerics software and non-parametric statistical analyses (SPSS V.17.0) were used. Observed differences in the predominant microbiota were subsequently confirmed and quantified with real-time PCR. Results Five bacterial species characterised dysbiosis in CD, namely a decrease in Dialister invisus (p¼0.04), an uncharacterised species of Clostridium cluster XIVa (p¼0.03), Faecalibacterium prausnitzii (p<1.33105) and Bifidobacterium adolescentis (p¼5.43106), and an increase in Ruminococcus gnavus (p¼2.13107). Unaffected relatives of patients with CD had less Collinsella aerofaciens (p¼0.004) and a member of the Escherichia colieShigella group (p¼0.01) and more Ruminococcus torques (p¼0.02) in their predominant microbiota as compared with healthy subjects. Conclusion Unaffected relatives of patients with CD have a different composition of their microbiota compared with healthy controls. This dysbiosis is not characterised by lack of butyrate producing-bacteria as observed in CD but suggests a role for microorganisms with mucin degradation capacity.

BACKGROUND AND AIMS The role of the commensal microbiota in the initiation and perpetuation of the intestinal inﬂammation in Crohn’s disease (CD) is well established.1e5 A general dysbiosis in patients with CD has been described with both culture-dependent and culture-independent techniques,6e8 but whether the observed dysbiosis is either cause or consequence of the disease remains unclear.4 Several studies have reported an overall lower biodiversity in patients with CD. At the groupspeciﬁc level, lower numbers of members of the Firmicutes phylum and of the Clostridium cluster IV combined with higher numbers of Bacteroides Gut 2011;60:631e637. doi:10.1136/gut.2010.223263

Significance of this study What is already known about this subject?

< The role of bacteria in the pathogenesis of Crohn’s

disease (CD) is well established (clinical presentation and lesions of the disease are localised in areas of highest bacterial exposure, clinical response of patients to diversion of the faecal stream and relapse on restoration or exposure to faecal material, clinical response to antibacterial treatment, absence of disease in germ-free animal models, susceptibility genes for CD are involved in recognition of and defence against microorganisms). < In patients, a general dysbiosis has been described with at the group-specific level lower numbers of members of the Firmicutes phylum and of the Clostridium cluster IV combined with higher numbers of Bacteroides. < Until now, the species-specific finding that has been most convincingly replicated is the lower presence of Faecalibacterium prausnitzii in patients with CD both in faecal samples and in biopsies.

What are the new findings? < We now provide the first detailed description of

this dysbiosis in patients with CD.

< Based on differences in the predominant faecal

microbiota in the largest cohort studied so far, we were able to describe a dysbiosis signature associated with CD, characterised by five bacterial species, namely Dialister invisus, an uncharacterised species of Clostridium cluster XIVa, Faecalibacterium prausnitzii, Bifidobacterium adolescentis and Ruminococcus gnavus. < This dysbiosis signature was markedly characteristic for the disease as it was not observed in unaffected relatives despite a common genetic background and shared nutritional habits. < We demonstrate further for the first time that unaffected relatives of patients with CD also have a different composition of their microbiota compared with healthy controls.

How might it impact on clinical practice in the foreseeable future? < This first detailed description of CD-associated dysbiosis will break new ground in unravelling the role of bacteria in CD. < Furthermore, uncovering this subclinical dysbiosis in unaffected subjects at risk allowed us to hypothesise that enhanced mucin degradation capacity of the intestinal microbiota might be an intermediate step towards CD and disease-associated dysbiosis. 631

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Inflammatory bowel disease have been reported.9e11 In several studies a proportional abundance of mucosa-associated Escherichia coli in CD has been described, but these ﬁndings are self-evidently limited to mucosal samples.12e14 Until now, the species-speciﬁc ﬁnding that has been most convincingly replicated is the lower presence of Faecalibacterium prausnitzii in patients with CD both in faecal samples and in biopsies.7 15e17 Moreover, the ability to produce butyrate and the anti-inﬂammatory properties of this species provide a coherent biological explanation for its lower presence in patients with CD.15 A more extended and detailed description of the observed dysbiosis in CD is, however, still lacking. The abnormal response to commensal intestinal bacteria in CD might be driven by the genetic background of the host, as many of the identiﬁed susceptibility genes for CD play a role in bacterial sensing. The ﬁrst gene identiﬁed in CD, NOD2/ CARD15, belongs to the family of pattern recognition receptors (PRRs) which are part of the human innate immunity.18 19 PRRs recognise speciﬁc bacterial sequences and are responsible for the defence against those organisms. Toll-like receptors (TLRs) were the ﬁrst class of cellular PRRs identiﬁed and are the best known PRRs.20 An association between the functional Asp299Gly polymorphism in TLR-4 and inﬂammatory bowel disease (IBD) has been demonstrated.21 Also the importance of autophagy in CD became clear as a result of genome-wide association scanning which has identiﬁed the autophagy-related 16-like 1 (ATG16L1) gene and immune-related guanosine triphosphatase (IRGM).22 23 To test the impact of the host genotype on the intestinal microbiota, comparisons were made between faecal samples from subjects with differing genetic relatedness, varying from unrelated persons to monozygotic twins.24e26 A higher similarity between samples was found with a higher degree of genetic relatedness. These studies thus indicated a signiﬁcant impact of the host genotype on the bacterial composition of faecal samples but did not ﬁnd a clear environmental effect. The interaction of host genotype with the composition of the intestinal microbiota is still poorly understood, but CD twin studies provide evidence that the bacterial composition is more inﬂuenced by disease status than by the genetic background of the host.7 17 Although supported by little evidence, the composition of the intestinal microbiota is also likely to be inﬂuenced by other factors such as dietary habits and the environment. Both genetic and environmental factors are shared within families, and ﬁrstdegree relatives of patients with CD are at much higher risk of developing CD as compared with the general population.27 In this regard, we previously reported a 57-fold increase in the incidence of IBD within multiply affected families.28 29 The aim of the current study was to investigate to what extent the predominant faecal microbiota of patients with CD has unique characteristics that discriminates them from healthy subjects with and/or without a shared familial background. Furthermore, we hypothesised that intestinal dysbiosis might be a precursor for CD and, as a consequence, that relatives, or a subgroup of them, also might have asymptomatic intestinal dysbiosis.

a speciﬁc diet, using probiotics or suffering from irritable bowel syndrome (IBS) were excluded from the analyses. A total of 68 patients with CD from 25 families, 84 of their unaffected relatives and 55 healthy subjects from 10 control families were studied (table 1). Four families had only one member affected with CD; 21 families had at least three affected members. Of the 84 unaffected relatives in our cohort, 80 were ﬁrst-degree (thus primary) relatives of patients with CD (48 siblings, 31 parents of patients with CD and one unaffected subject that had both affected siblings and offspring). The four that did not have ﬁrstdegree relatives with CD lived in a close relationship with four patients with CD. We used large families which resulted overall in balanced numbers of unaffected relatives and patients per family. The median number of unaffected relatives used per family was 3 (IQR 1e4). For patients, the median number per family was also 3 (IQR 2e3). The median ratio of unaffected relatives to patients per family was 1 (IQR 0.67e1.75). All patients were in clinical remission at the time of sampling. The patient characteristics at the time of sampling are summarised in table 2. The Ethics Committee of the University of Leuven (Belgium) approved the study, and all participants gave informed consent. Faecal samples were collected from each subject and upon collection they were immediately frozen at 808C and stored for DNA extraction.

Denaturing gradient gel electrophoresis (DGGE) profiling To identify potential microbial characteristics of CD, an objective comparison by means of a community ﬁngerprinting approach based on DGGE analysis of ribosomal PCR amplicons was performed.30 Therefore, total bacterial DNA was extracted from the faecal samples using the method of Pitcher and colleagues with slight modiﬁcations.31 Next, community PCR was conducted using universal primers F357+GC clamp and R518 targeting the hypervariable V3 region of the 16S rRNA gene. The resulting 16S rRNA gene amplicons were analysed with DGGE using a 35e70% gradient as previously described,31 resulting in proﬁles of the predominant faecal microbiota per subject. On each DGGE gel, a standard reference consisting of an amplicon mix of 12 different bacterial species was included in the middle and at both outer ends for digital gel normalisation and to allow comparison between gels.

Digital processing of DGGE profiles DGGE proﬁles were digitally processed with BioNumerics software version 4.6 (Applied Maths, St-Martens-Latem, Belgium) in a multistep procedure following the manufacturer ’s instructions. After normalisation of the gels, individual bands in each sample lane were marked using the auto search bands option, followed by manual correction if necessary. All proﬁles were compared using the band matching tool and uncertain bands were included in the position tolerance settings. Allocation of bands to so-called band-classes was performed automatically Table 1

Faecal samples used for analysis represented per group

METHODS Subjects and study design Microbial DNA from stool samples of patients with CD was compared with that of their unaffected relatives and healthy subjects from families of comparable generation, gender distribution and family composition for whom the recruitment and matching criteria were previously described.29 Subjects taking antibiotics within 4 weeks before the faecal sampling, following 632

Age (years)* Gender (% females) Age at diagnosis (years)* Disease duration (years)*

Patients with CD n[68

Unaffected relatives n[84

Control subjects n[ 55

45 (25e76) 59% 28 (12e55) 17 (2e57)

52 (14e86) 50% NA NA

50 (28e78) 55% NA NA

*The mean is given, with the minimum and maximum in parentheses. CD, Crohn’s disease; NA, not applicable.

Gut 2011;60:631e637. doi:10.1136/gut.2010.223263

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Inflammatory bowel disease Table 2

Patient characteristics at the time of sampling No. of patients with CD

Location* Ileum (L1) Colon (L2) Ileocolon (L3) Behaviour* Non-stricturing, non-penetrating (B1) Structuring (B2) Penetrating (B3) Previous surgery* Yes No Medical treatment at time of sampling* Immunosuppressants Steroids NSAIDs Sulfasazines Mesalazines

27 6 31 25 22 13 40 20 10 3 3 12 8

*Maximal disease extent at the time of sampling was available in 64/68 patients; disease behaviour at last follow-up before the time of sampling was available in 60/68 patients; previous surgery at the time of sampling was available in 60/68 patients (type of surgery: 5/40 unknown, 1/40 gastric Roux-en-Y anastomosis, 16/40 partial ileal and partial colonic resection, 2/40 (sub)total colectomies, 11/40 hemicolectomies, 5/40 partial ileal resection); medical treatment at the time of sampling was available in 57/68 patients. CD, Crohn’s disease; NSAIDs, non-steroidal anti-inflammatory drugs.

and again checked manually. Essentially, band-classes were arbitrarily generated in a collective analysis of all proﬁles by tracing common bands across different sample proﬁles. This allowed every band in a proﬁle to be assigned to its nearest bandclass. A maximum error of 0.5% deviation was applied, which means that a band was only allocated if it was located at a distance of <0.5% of the total length of the proﬁle from the closest band-class. The designation of the band-classes was based on their position on the proﬁle compared with the standard reference. The intensity of a given band-class was expressed with respect to the other band-classes on the same proﬁle.

Comparative analyses of the DGGE profiles The quantitative information derived from relative band intensities per subject and per band-class was exported as a data matrix. Bacterial groups present in quantities under the detection limit were denoted as zero. Non-parametric tests were applied to compare differences in the complex proﬁles across groups (table 1). Therefore ManneWhitney U tests on the intensity per band-class were used to compare the groups in SPSS V.17.0. To correct for multiple testing, all p values were multiplied by the number of band-classes obtained (n¼75). Only two-tailed p values <0.05 after correction were considered signiﬁcant.

DGGE band extraction and sequencing DGGE bands that differed signiﬁcantly between groups according to the statistical analysis were puriﬁed from the complex faecal samples, before they were sequenced, by excision from the gel. The DNA was eluted from the gel slice into 15 ml of TE buffer by heating the mixture of buffer and gel for 10 min at 658C. The resulting DNA solution was then ampliﬁed again using the universal V3 16S rRNA primers F357+GC clamp and R518. Puriﬁcation of the bands was facilitated by using DGGE gel expansion where the amplicons were checked by DGGE with an adjusted gradient32 and again excised at least three times until a single band was obtained. The purpose of repeating DGGE until one single band is obtained is only to purify the Gut 2011;60:631e637. doi:10.1136/gut.2010.223263

PCR product before it can be sequenced. In a ﬁnal DGGE round, puriﬁed amplicons were analysed together with the original sample proﬁles from which they were excised to check visually whether the correct bands were puriﬁed. When puriﬁed bands matched with the targeted bands in the sample proﬁles, amplicons were selected for sequencing using an ABI Prism 3130 Genetic Analyzer (Applied Biosystems, Foster City, California, USA). Sequences were read from both directions with primers F357 (without clamp) and R518, respectively. The sequence was assembled with BioNumerics software. Homology searches of the GenBank DNA database were performed with BLAST search. Based on the BLAST results, reference sequences of phylogenetic neighbour species (up to 90% similarity) were included for clustering analysis using multiple sequence alignment and the average linking method to conﬁrm allocation of the puriﬁed band sequences to the most probable species.

Real-time PCR Bacterial species that differed signiﬁcantly in the comparative DGGE band-class analyses were quantiﬁed with the Applied Biosystems 7500 Fast Real-Time PCR (RT-PCR) System in the fast 7500 mode. Primers (Primer express 3.0 software) that spanned a target region in the V3 region of the 16S rRNA gene species of interest were selected (Supplementary table 1). The RT-PCR was carried out in a 20 ml total reaction mixture using 4 ml of DNA. Calibration curves were constructed using dilutions of genomic DNA from a control strain (Supplementary table 1) for which the number of bacteria was determined by plate counting. The quantitative results of the RT-PCR analyses were based on measurements in triplicate with a maximum variation <0.5 Ct. Amounts of bacterial species were expressed as log10 values per gram wet weight of faeces.

RESULTS Differences in five bacterial species characterise predominant dysbiosis in CD In the 207 PCR-DGGE proﬁles analysed in this study, 75 bandclasses were identiﬁed. We ﬁrst compared patients with CD versus all unaffected controls (unaffected relatives and unrelated controls taken together) in search of disease-associated characteristics of the predominant faecal microbiota. After Bonferroni correction for the 75 band-classes, signiﬁcant differences were observed between the patients with CD and unaffected subjects for seven band-classes. To assign these seven band-classes to a bacterial species or phylotype, DNA was puriﬁed per bandclass from at least ﬁve different samples originating from different patients, unaffected relatives and controls. Puriﬁed amplicons were analysed together with the original sample proﬁles from which they were excised, to check visually whether the correct bands were puriﬁed. After puriﬁcation, band-class 3.72 appeared to be composed of different bands. We thus concluded that this band-class did not represent a single bacterial species and therefore it was not studied further. All other puriﬁed bands of the remaining six band-classes matched with the targeted bands in the sample proﬁles, and these amplicons were sequenced. Within each band-class all generated sequences (length 135 bp, representing 8.7e9.5% of the 16S rRNA gene) were identical. Band-classes were tentatively assigned to a bacterial species or phylotype based on the highest (98e100%) sequence similarity match to GenBank sequences obtained by BLAST analysis (Supplementary table 2). These species assignments were subsequently conﬁrmed by clustering analysis with reference sequences of the closest phylogenetic neighbours. 633

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Inflammatory bowel disease Figure 1 Examples of denaturing gradient gel electrophoresis (DGGE) profiles. The letters next to the picture correspond to the letters in the right table per bandclass. Each of the six lanes shown on the left represents a DGGE profile visualising the predominant faecal microbiota of one subject.

the left picture.

In this way, band-classes 9.68 and 11.51 were both assigned to F prausnitzii, band-class 11.00 was assigned to Ruminococcus gnavus, band-class 14.60 was assigned to Dialister invisus and band-class 16.24 was assigned to Biﬁdobacterium adolescentis. Based on its phylogenetic position revealed by clustering analysis, band-class 10.79 could be assigned to an uncharacterised species of Clostridium cluster XIVa. The disproportion of these ﬁve bacterial species that characterise the observed differences between the patients with CD and unaffected subjects are represented per band-class in ﬁgure 1 and in ﬁgure 2 the occurrence in the DGGE proﬁles is expressed in terms of a percentage. We also analysed the predominant faecal proﬁles of the patients with CD according to the clinical phenotypes described in table 2. However, none of the dysbalanced bacterial species was associated with a speciﬁc clinical phenotype. Only when comparing predominant microbial proﬁles of the subgroup of patients with CD that underwent an ileocolonic resection (n¼16) prior to sampling with the subgroup of patients that never had an intestinal operation (n¼20) was a signiﬁcant quantitative difference in F prausnitzii revealed (corrected p value¼0.002) with even less F prausnitzii in patients with CD after ileocolonic resection.

Subclinical dysbiosis in asymptomatic relatives of patients with CD Proﬁles of unaffected relatives were compared with those of unrelated control subjects in search of asymptomatic intestinal dysbiosis in relatives. Three band-classes differed signiﬁcantly between both unaffected groupsdthat is, band-classes 8.55, 11.35 and 16.89 (table 3). Interestingly, the differences observed between unaffected relatives and unrelated control subjects were also observed in patients with CD compared with unrelated controls, but to a lesser extent, and therefore were not signiﬁcantly different. For the predominant occurrence of the bacterial species that (statistically) differed between unaffected relatives of patients with CD and healthy controls, the values of the unaffected relatives are at one extreme, those of the healthy controls at the other extreme, and those of patients with CD are in between. The prevalence of patients with CD for these bandclasses was thus situated in between the prevalence of unrelated

Influence of the genetic and environmental background on the predominant microbiota Because the composition of the intestinal microbiota is likely to be inﬂuenced by many factors, and both genetic and environmental factors are shared within families, we next compared patients with their unaffected relatives. No differences could be found between patients or unaffected relatives according to the number of affected relatives. In other words, we did not ﬁnd differences between single incidence families and multiplex families. Therefore, we combined both groups. However, when comparing the patient group only with their unaffected relatives, all six band-classes except band-class 10.79 (unknown species of Clostridium cluster XIVa) remained signiﬁcantly different between both groups and no additional discriminative band-classes were identiﬁed. To evaluate the effect of the genetic and environmental background on the observed differences in patients with CD, we then compared the patient group with unrelated control subjects only, to allow a larger potential impact of genetic and environmental factors. All except band-classes 11.51 (F prausnitzii) and 14.60 (D invisus) remained signiﬁcantly different between both unrelated groups. 634

Figure 2 Denaturing gradient gel electrophoresis (DGGE) band-classes that differed in the predominant faecal microbiota of patients with Crohn’s disease (CD), their unaffected relatives and control subjects, expressed as a percentage of subjects with the respective band in their predominant profile per band-class and per group. Inclined, the corresponding bacterial species are added. Gut 2011;60:631e637. doi:10.1136/gut.2010.223263

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Inflammatory bowel disease Table 3 Median intensity of band-classes that differed significantly between unaffected relatives of patients with Crohn’s disease and unrelated controls Median intensity (IQR) Bandclass 8.55 11.35 16.89

Bacterial species R torques Unspecified member of the Escherichia colieShigella group C aerofaciens

Unaffected relatives n[84

Unrelated controls n[55

Corrected p value

1.83 (0.42e3.74) 0 (0e0)

0 (0e2.21) 0 (0e0.20)

0.02 0.01

0 (0e0)

0 (0e0.78)

0.004

controls and that of unaffected relatives (ﬁgure 2). After puriﬁcation and sequencing, BLAST analysis in GenBank revealed 100% similarity of band-class 8.55 to Coprococcus comes, Clostridium nexile and Ruminococcus torques. Band-class 11.35 showed 100% sequence similarity to members of the E colieShigella group whereas band-class 16.89 was identiﬁed as Collinsella aerofaciens. Subsequent clustering analysis conﬁrmed that bandclass 11.35 can be assigned to a member of the E colieShigella group, and that band-class 16.89 could be assigned to C aerofaciens. Clustering analysis indicated that R torques rather than C comes or C nexile was phylogenetically the closest species corresponding to band-class 8.55. To check for family-associated effects on band-class distribution, we analysed the band-classes using family as the grouping variable and did not ﬁnd signiﬁcant differences for the bandclasses that we identiﬁed as characterising CD-associated dysbiosis or for those characterising subclinical dysbiosis in unaffected relatives.

Quantitative validation of disease-associated differences in predominant microbiota with RT-PCR CD-associated predominant dysbiosis was consistently determined by differences in band-classes assigned to F prausnitzii, R gnavus and B adolescentis in the different comparisons made. Therefore, these three species were quantiﬁed with RT-PCR. As the quantitative data obtained were not normally distributed (KolmogoroveSmirnova test statistics: p values <0.023), we performed a non-parametric ManneWhitney U test to compare the groups. The median number of F prausnitzii was 9.58 log10/g in patients with CD (IQR 7.40e11.07 log10/g) and 11.45 log10/g in unaffected controls (IQR 10.92e11.90 log10/g) (p¼2.6131011) (ﬁgure 3A). The median number of R gnavus was 8.44 log10/g in patients (IQR 7.17e9.51 log10/g) and 7.51 log10/g in unaffected controls (IQR 6.31e8.81 log10/g) (p¼6.003103) (ﬁgure 3B). The median number of B adolescentis

Figure 3

was 5.00 log10/g in patients (IQR 4.31e6.65 log10/g) and 8.34 log10/g in unaffected controls (IQR 5.21e9.14 log10/g) (p¼1.5931010) (ﬁgure 3C). The differences in the quantitative data per group reﬂect the disproportions identiﬁed in the predominant microbiota and provide a quantitative validation of the DGGE band-class analysis results.

DISCUSSION In this study we identiﬁed ﬁve disproportionate bacterial species that characterise dysbiosis in the predominant faecal microbiota of a large cohort of patients with CD. This dysbiosis signature we found in patients with CD was markedly characteristic for the disease as it was not observed in unaffected relatives despite a common genetic background and shared nutritional habits. The butyrate producer F prausnitzii occurred signiﬁcantly less in the predominant microbial proﬁles of patients with CD. In agreement with recent ﬁndings of Sokol and co-workers, signiﬁcantly lower numbers of this species were also detected by RT-PCR in patients with CD as compared with unaffected subjects.33 In contrast, we found that R gnavus appeared more in the predominant faecal microbiota of patients with CD. The higher occurrence of R gnavus in patients with CD has been described previously in mucosal samples of patients with ileal disease.34 We were also able to demonstrate the higher prevalence of this species in faecal samples of patients with CD, both in the predominant proﬁles and by quantitative PCR. In vitro, R gnavus expresses very high b-glucuronidase activity.35 b-Glucuronidase activity can induce the formation of toxic compounds in the colon which might cause or perpetuate local inﬂammation. A third bacterial species that clearly differed between patients with CD and unaffected controls was B adolescentis. The lower occurrence of biﬁdobacteria in general in patients with CD has been described earlier,36 but we have now speciﬁcally demonstrated for the ﬁrst time the lower prevalence of B adolescentis both qualitatively and quantitatively. To our knowledge, the lower occurrence of D invisus in patients with CD has never been reported before. Interestingly, this species was ﬁrst isolated from dental root infections,37 but has also been detected as part of the normal intestinal microbiota.38 The sequence of band-class 10.79 was assigned to an uncharacterised species of Clostridium cluster XIVa that was under-represented in the predominant microbial proﬁles of patients with CD. This sequence showed 100% similarity with the 16S rRNA sequence of strain SS2/1 which was previously found in the colonic microbiota of three healthy adult humans and was classiﬁed as a Clostridium indolis-like member of Clostridium cluster XIVa.39 Although this unknown species has

Boxplots of the number of (A) F. prausnitzii (B) R. gnavus and (C) B. adolescentis in log10 per g of faeces per group. CD, Crohn’s disease.

Gut 2011;60:631e637. doi:10.1136/gut.2010.223263

635

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Inflammatory bowel disease not been fully taxonomically characterised, its clear phylogenetic afﬁliation to butyrate producers of cluster XIV and its lower occurrence in patients with CD can be considered analogous to the observations for F prausnitzii. So far, however, the lower prevalence of this unknown Clostridium species has not been related to CD. Another major ﬁnding in this study is that in unaffected relatives asymptomatic differences in three other bacterial species were also identiﬁed. Both an unspeciﬁed member of the E colieShigella group and C aerofaciens were less prevalent in relatives of patients with CD. Strains of C aerofaciens were previously isolated from patients with colon cancer, ulcerative colitis and CD.40 The lower prevalence of this species in unaffected relatives compared with unrelated controls is a novel ﬁnding. Band-class 8.55, which was found more in relatives of patients with CD, phylogenetically most closely resembled C nexile, R torques and C comes. The latter two commensal species have previously been linked to CD.16 41 A possible role for C comes in the pathogenesis of CD has been suggested based on its interaction with the immune system.42 C comes can activate complement and thereby induce inﬂammation. Moreover, these bacteria are not ingested by neutrophils and can bind immunoglobulins through their Fc region which prevents phagocytosis of this microorganism.42 For R. torques no role in CD pathophysiology has been suggested so far. Like R gnavus, R torques belongs to Clostridium cluster XIVa, but, in contrast to the other members of this group, R gnavus and R torques are non-butyrate-producing members.43 Moreover, like R gnavus and Candida albicans, R torques is known to degrade gastrointestinal mucin.43 44 For C albicans, the capacity to break down mucus and penetrate the mucin barrier also plays a role in its virulence.44 Degradation of the mucosal barrier might thus enhance bacterial translocation and lead to increased permeability of the gut. C albicans was more frequently isolated from stool samples from patients with CD (44%) and their unaffected relatives (38%) compared with controls (22%) in a cohort comprising similar French families and the same subjects used for the present study.45 As R torques was also more frequently found in relatives of patients with CD, an enhanced mucin degradation capacity of the intestinal microbiota might precede or predispose further to CD. The intestinal microbiota of patients with CD has been studied with different molecular techniques. Several ﬁndings had previously been reported in various publications, but so far these were only partial and/or not very persuasive. In this study DGGE was used as a molecular technique to explore predominant differences in faecal samples of a large cohort of patients with CD, their unaffected relatives and healthy controls. We were able to describe in detail differences among the most abundant members of the faecal microbiota and to conﬁrm and integrate previous ﬁndings. In our opinion, large cohorts are needed to study complex communities and, despite being laborious, DGGE is a robust technique and highly valuable as an exploratory tool. The recruitment of the patients with CD through a tertiary referral centre resulted in a cohort of patients with complicated CD. Moreover, the speciﬁc design of the study, using both related controls at increased risk for CD (mainly from multiplex families) and independent control subjects, precluded us from also powering the study for potential effects of disease phenotype or treatment. Subanalyses only revealed that patients with CD might have even less faecal F prausnitzii after ileocolonic resection compared with those that were never operated on. 636

As the group of patients in which we could evaluate this was very small, whether this is a true ﬁnding should be studied further. Despite the potential bias using this selected cohort focusing on multiplex families, the CD-associated predominant dysbiosis we found is in agreement with previous ﬁndings. In conclusion, we provide for the ﬁrst time a set of ﬁve bacterial species that characterise the predominant dysbiosis in CD. We identiﬁed a faecal microbiota dysbiosis signature associated with CD, characterised by a decreased presence of F prausnitzii, B adolescentis, D invisus and an unknown species of Clostridium cluster XIVa, and an increased presence of R gnavus. The disproportions of three of these predominant species were also quantitatively conﬁrmed, and the identiﬁcation suggests lack of butyrate-producing capacity in conjunction with mucin degradation as plausible physiopathological explanations. Furthermore, the novelty in the experimental approach is that not only were the predominant microbiota of patients with CD and healthy controls studied, but for the ﬁrst time comparisons with a large cohort of unaffected relatives of patients with CD were also included. In contrast to CD-associated dysbiosis, the dysbiosis in relatives of patients was not characterised by a lack of butyrate-producing capacity, but similar enhanced mucin degradation could be hypothesised. As the mucosal barrier is the ﬁrst-line defence of the host against intestinal bacteria, the shift from normobiosis to the dysbiosis observed in relatives of patients with CD might be an intermediate step towards CD and disease-associated dysbiosis. Despite the fact that we did not study the overall butyrate-producing or mucin degradation capacity of the microbiota in this cohort, it is intriguing to ﬁnd a functional overlap between dysbalanced bacteria in patients with CD and their unaffected relatives at risk. Further investigation should reveal not only if the balance between butyrate production and mucin degradation is indeed essential in predisposing to CD but also how this balance can be inﬂuenced. Funding The project was funded by the Fund for Scientific Research-Flanders, Belgium (FWO-Vlaanderen), research project G.0455.06N. GH, SV and VDP are postdoctoral fellows of the FWO-Vlaanderen. Competing interests None. Ethics approval This study was conducted with the approval of the ethics committee of the Catholic University of Leuven. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

Rutgeerts P, Geboes K, Peeters M, et al. Effect of fecal stream diversion on recurrence of Crohns-disease in the neoterminal ileum. Lancet 1991;338:771e4. D’Haens GR, Geboes K, Peeters M, et al. Early lesions of recurrent Crohn’s disease caused by infusion of intestinal contents in excluded ileum. Gastroenterology 1998;114:262e7. Sartor RB. Microbial influences in inflammatory bowel diseases. Gastroenterology 2008;134:577e94. Seksik P, Sokol H, Lepage P, et al. Review article: the role of bacteria in onset and perpetuation of inflammatory bowel disease. Aliment Pharmacol Ther 2006;24:11e18. Marteau P, Lepage P, Mangin I, et al. Review article: gut flora and inflammatory bowel disease. Aliment Pharmacol Ther 2004;20:18e23. Frank DN, Amand ALS, Feldman RA, et al. Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases. Proc Natl Acad Sci USA 2007;104:13780e5. Willing B, Halfvarson J, Dicksved J, et al. Twin studies reveal specific imbalances in the mucosa-associated microbiota of patients with ileal Crohn’s disease. Inflamm Bowel Dis 2009;15:653e60. Sokol H, Seksik P, Rigottier-Gois L, et al. Specificities of the fecal microbiota in inflammatory bowel disease. Inflamm Bowel Dis 2006;12:106e11. Scanlan PD, Shanahan F, O’Mahony C, et al. Culture-independent analyses of temporal variation of the dominant fecal microbiota and targeted bacterial subgroups in Crohn’s disease. J Clin Microbiol 2006;44:3980e8.

Gut 2011;60:631e637. doi:10.1136/gut.2010.223263

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Inflammatory bowel disease 10. 11. 12. 13. 14. 15. 16.

17. 18. 19. 20. 21.

22. 23. 24. 25. 26. 27.

Andoh A, Tsujikawa T, Sasaki M, et al. Faecal microbiota profile of Crohn’s disease determined by terminal restriction fragment length polymorphism analysis. Aliment Pharmacol Ther 2009;29:75e82. Manichanh C, Rigottier-Gois L, Bonnaud E, et al. Reduced diversity of faecal microbiota in Crohn’s disease revealed by a metagenomic approach. Gut 2006;55:205e11. Darfeuille-Michaud A, Neut C, Barnich N, et al. Presence of adherent Escherichia coli strains in ileal mucosa of patients with Crohn’s disease. Gastroenterology 1998;115:1405e13. Martin HM, Campbell BJ, Hart CA, et al. Enhanced Escherichia coli adherence and invasion in Crohn’s disease and colon cancer. Gastroenterology 2004;127:80e93. Sasaki M, Sitaraman SV, Babbin BA, et al. Invasive Escherichia coli are a feature of Crohn’s disease. Lab Invest 2007;87:1042e54. Sokol H, Pigneur B, Watterlot L, et al. Faecalibacterium prausnitzii is an anti-inflammatory commensal bacterium identified by gut microbiota analysis of Crohn disease patients. Proc Natl Acad Sci USA 2008;105:16731e6. Martinez-Medina M, Aldeguer X, Gonzalez-Huix F, et al. Abnormal microbiota composition in the ileocolonic mucosa of Crohn’s disease patients as revealed by polymerase chain reaction-denaturing gradient gel electrophoresis. Inflamm Bowel Dis 2006;12:1136e45. Dicksved J, Halfvarson J, Rosenquist M, et al. Molecular analysis of the gut microbiota of identical twins with Crohn’s disease. ISME J 2008;2:716e27. Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucine-rich repeat variants with susceptibility to Crohn’s disease. Nature 2001;411:599e603. Ogura Y, Bonen DK, Inohara N, et al. A frameshift mutation in NOD2 associated with susceptibility to Crohn’s disease. Nature 2001;411:603e6. Medzhitov R. Toll-like receptors in innate and adaptive immunity. Mol Biol Cell 2000;11:282A. Franchimont D, Vermeire S, El Housni H, et al. Deficient hostebacteria interactions in inflammatory bowel disease? The toll-like receptor (TLR)-4 Asp299gly polymorphism is associated with Crohn’s disease and ulcerative colitis. Gut 2004;53:987e92. Hampe J, Franke A, Rosenstiel P, et al. A genome-wide association scan of nonsynonymous SNPs identifies a susceptibility variant for Crohn disease in ATG16L1. Nat Genet 2007;39:207e11. Parkes M, Barrett JC, Prescott NJ, et al. Sequence variants in the autophagy gene IRGM and multiple other replicating loci contribute to Crohn’s disease susceptibility. Nat Genet 2007;39:830e2. Vandemerwe JP, Stegeman JH, Hazenberg MP. The resident fecal flora is determined by genetic-characteristics of the hostdimplications for Crohn’s disease. Antonie Van Leeuwenhoek 1983;49:119e24. Zoetendal EG, Akkermans ADL, Vliet WMA, et al. The host genotype affects the bacterial community in the human gastronintestinal tract. Microb Ecol Health Dis 2001;13:129e34. Stewart JA, Chadwick VS, Murray A. Investigations into the influence of host genetics on the predominant eubacteria in the faecal microflora of children. J Med Microbiol 2005;54:1239e42. Russell RK, Satsangi J. IBD: a family affair. Best Pract Res Clin Gastroenterol 2004;18:525e39.

Gut 2011;60:631e637. doi:10.1136/gut.2010.223263

28. 29. 30.

31.

32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.

Joossens M, Van Steen K, Branche J, et al. Familial aggregation and antimicrobial response dose-dependently affect the risk for Crohn’s disease. Inflamm Bowel Dis 2010;16:58e67. Van Kruiningen HJ, Joossens M, Vermeire S, et al. Environmental factors in familial Crohn’s disease in Belgium. Inflamm Bowel Dis 2005;11:360e5. Huys G, Vanhoutte T, Vandamme P. Application of sequence-dependent electrophoresis fingerprinting in exploring biodiversity and population dynamics of human intestinal microbiota: what can be revealed? Interdiscip Perspect Infect Dis 2008;2008:597e603. Vanhoutte T, Huys G, De Brandt E, et al. Temporal stability analysis of the microbiota in human feces by denaturing gradient gel electrophoresis using universal and group-specific 16S rRNA gene primers. FEMS Microbiol Ecol 2004;48:437e46. Gafan GP, Spratt DA. Denaturing gradient gel electrophoresis gel expansion (DGGEGE)dan attempt to resolve the limitations of co-migration in the DGGE of complex polymicrobial communities. FEMS Microbiol Lett 2005;253:303e7. Sokol H, Seksik P, Furet JP, et al. Low counts of Faecalibacterium prausnitzii in colitis microbiota. Inflamm Bowel Dis 2009;15:1183e9. Prindiville T, Cantrell M, Wilson KH. Ribosomal DNA sequence analysis of mucosa-associated bacteria in Crohn’s disease. Inflamm Bowel Dis 2004;10:824e33. Beaud D, Tailliez P, Anba-Mondoloni J. Genetic characterization of the betaglucuronidase enzyme from a human intestinal bacterium, Ruminococcus gnavus. Microbiology 2005;151:2323e30. Favier C, Neut C, Mizon C, et al. Fecal beta-D-galactosidase production and Bifidobacteria are decreased in Crohn’s disease. Dig Dis Sci 1997;42:817e22. Downes J, Munson M, Wade WG. Dialister invisus sp nov., isolated from the human oral cavity. Int J Syst Evol Microbiol 2003;53:1937e40. Rajilic-Stojanovic M, Smidt H, de Vos WM. Diversity of the human gastrointestinal tract microbiota revisited. Environ Microbiol 2007;9:2125e36. Louis P, Duncan SH, Mccrae SI, et al. Restricted distribution of the butyrate kinase pathway among butyrate-producing bacteria from the human colon. J Bacteriol 2004;186:2099e106. Kageyama A, Benno Y, Nakase T. Phylogenetic and phenotypic evidence for the transfer of Eubacterium aerofaciens to the genus Collinsella as Collinsella aerofaciens gen. nov., comb. nov. Int J Syst Bacteriol 1999;49:557e65. Vandemerwe JP, Schroder AM, Wensinck F, et al. The obligate anaerobic faecal flora of patients with Crohn’s disease and their first-degree relatives. Scand J Gastroenterol 1988;23:1125e31. Vandemerwe JP, Stegeman JH. Binding of Coprococcus comes to the Fc portion of IgG. A possible role in the pathogenesis of Crohn’s disease? Eur J Immunol 1985;15:860e3. Dethlefsen L, Eckburg PB, Bik EM, et al. Assembly of the human intestinal microbiota. Trends Ecol Evol 2006;21:517e23. Colina AR, Aumont F, Deslauriers N, et al. Evidence for degradation of gastrointestinal mucin by Candida albicans secretory aspartyl proteinase. Infect Immun 1996;64:4514e19. Standaert-Vitse A, Chamaillard M, Joossens M, et al. Candida albicans colonization and ASCA in Familial Crohn’s Disease. Am J Gastroenterol 2009;104:1745e53.

637

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology

Neuronal stimulation with 5-hydroxytryptamine 4 receptor induces anti-inflammatory actions via a7nACh receptors on muscularis macrophages associated with postoperative ileus Yasuaki Tsuchida,1,2 Fumihiko Hatao,2 Masahiko Fujisawa,3 Takahisa Murata,1 Michio Kaminishi,2 Yasuyuki Seto,2 Masatoshi Hori,1 Hiroshi Ozaki1 < Additional figures are

published online only. To view these files please visit the journal online (http://gut.bmj. com). 1

Department of Veterinary Pharmacology, Graduate School of Agriculture and Life Sciences, The University of Tokyo, Tokyo, Japan 2 Department of Metabolic Care and Gastrointestinal Surgery Graduate School of Medicine, The University of Tokyo, Tokyo, Japan 3 Department of Veterinary Science, Nippon Veterinary and Life Science University, Tokyo, Japan Correspondence to Dr Masatoshi Hori, Department of Veterinary Pharmacology, The University of Tokyo, 1-1-1 Yayoi, Bunkyo-ku, Tokyo 113-8657, Japan; [email protected] Revised 21 October 2010 Accepted 25 October 2010 Published Online First 29 November 2010

This paper is freely available online under the BMJ Journals unlocked scheme, see http:// gut.bmj.com/site/about/ unlocked.xhtml 638

ABSTRACT Background The main symptom of postoperative ileus (POI) is an intestinal motility disorder in which monocytes/macrophages and neutrophils play crucial roles. Prokinetic 5-hydroxytryptamine 4 receptor (5-HT4R) agonists and dopamine receptor antagonists are potential therapeutic agents for directly ameliorating the motility disorder associated with POI. Aim To determine the effects of the 5-HT4R agonists mosapride citrate (MOS) and CJ-033466 on intestinal smooth muscle contractility relative to immune reactions after POI. Methods Intestinal manipulation (IM) was applied to the rat distal ileum. Both MOS (0.3 and 1 mg/kg, s.c.) and CJ-033466 (1 mg/kg, s.c.) were administered to the animals before and after IM. At 24 h after IM, isolated intestinal smooth muscle contractile activity in vitro, gastrointestinal transit in vivo, inflammatory mediator expression and leucocyte infiltration were measured. Results After IM, ileal circular muscle contractility in vitro and gastrointestinal transit in vivo were reduced and the number of macrophages and neutrophils increased in the inflamed muscle layer, resulting in the induction of inflammatory mediators such as interleukin 1 b (IL-1b), IL-6, tumour necrosis factor a (TNFa), monocyte chemoattractant protein 1 (MCP-1) and inducible nitric oxide synthase (iNOS). Both MOS and CJ-033466 significantly attenuated not only the intestinal motility dysfunction but also the leucocyte infiltration and inflammatory mediator expression after IM. The autonomic ganglionic blocker hexamethonium (1 mg/kg, i.p.) and the a7-nicotinic acetylcholine receptor (a7nAChR) antagonist methyl lycaconitine citrate (0.087 mg/kg, i.p.) blocked MOS-mediated ameliorative actions. Immunohistochemically, a7nAChR is expressed by monocytes/macrophages but not by neutrophils in the inflamed intestine. Conclusion Stimulating the 5-HT4R accelerates acetyl choline (ACh) release from cholinergic myenteric neurons, which subsequently activates a7nAChR on activated monocytes/macrophages to inhibit their inflammatory reactions in the muscle layer. In addition to their gastroprokinetic action, 5-HT4R agonists might serve as novel therapeutic agents for POI characterised by anti-inflammatory potency.

Significance of this study What is already known about this subject?

< The 5-HT4R agonist mosapride is actually used

in clinical practice as a gastrointestinal prokinetic drug. < Recent clinical trials showed that mosapride, that is benzamide analogues of cisapride, reduce postoperative ileus although pharmacological mechanisms are not well understood. < Stimulation of the vagus nerve prevents postoperative ileus through a7nAChR in macrophages.

What are the new findings?

< The 5-HT4R agonist mosapride dramatically

inhibits postoperative ileus through anti-inflammatory reaction in addition to its gastrointestinal prokinetic action. < The anti-inflammatory action is mediated by acetylcholine (ACh) release from cholinergic myenteric neurons, which subsequently activates a7nAChR on activated monocytes/macrophages to inhibit inflammation. < For the first time, we found a new subset of activated macrophages expressed a7nAChR during inflammation in small intestine.

How might it impact on clinical practice in the foreseeable future? < 5-HT4R agonists, such as mosapride, could be clinically used not only as a gastrointestinal prokinetic drug but also as anti-inflammatory drug to prevent postoperative ileus. < Intestinal macrophages are unique immunoreactive cells inducing a7nAChR during inflammation, which give a significant impact on medical sciences. < 5-HT4R could be a new therapeutic molecular target for anti-inflammatory gastrointestinal diseases.

INTRODUCTION Various 5-HT4R agonists, such as tegaserod, cisapride and mosapride, have been clinically validated as treatment for gastrointestinal disorders characterised by dysmotility. Originally, 5-HT4R agonists were believed to induce prokinetic potency Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology only in the upper part of the gastrointestinal tract. Therefore, 5-HT4R agonists are still clinically used to treat gastroparesis and functional dyspepsia.1e3 However, 5-HT4R agonists can also induce prokinetic ability in the lower part of the gastrointestinal tract in experimental animals and humans,4 5 and they are clinically applied to treat chronic constipation and constipationpredominant irritable bowel syndrome.6 7 Indeed, immunohistochemical studies have identiﬁed 5-HT4R in the myenteric and submucosal ganglia of gastrointestinal tissues.8e10 Postoperative ileus (POI) is a common complication after intra-abdominal surgery that is accompanied by increased morbidity and prolonged hospitalisation, increasing hospital costs.11 Neurogenic, inﬂammatory and inﬂammatoryeneuronal interactive mechanisms are generally considered to induce POI.12 Sympathetic reﬂexes, the activation of non-adrenergic, non-cholinergic nerves, inhibitory humoural agents and the inﬂuence of anaesthetics play crucial roles in the pathogenesis of POI at the acute stage during and shortly after (up to 3 h) laparotomy and abdominal surgery.13 14 Local inﬂammatory responses in the manipulated intestine additionally participate in disordered motility during the later stage of POI (24 h after surgery).15e19 Many resident macrophages are distributed in the subserosal, myenteric plexus regions of the intestinal muscle layer and inside the circular smooth muscle layer under healthy conditions.20 21 Although these ramiﬁed forms of resident macrophages normally remain inactive,22 23 inﬂammatory stimuli and/or mechanical stress induced by IM activates them to produce prostaglandins, nitric oxides, inﬂammatory cytokines and chemokines that cause muscularis inﬂammation and intestinal motility disorders.15e19 24 The pharmacological management of POI is important to inhibit morbidity rates and reduce hospital costs and length of hospital stay. Gastrointestinal prokinetic agents might be useful for managing or preventing POI according to some clinical trials.25e28 The effects of various gastrointestinal prokinetic agents on postoperative ileus have also been investigated in experimental animals.29 Peripherally acting m-opioid receptor antagonists, such as alvimopan and methylnaltrexone, offer promising results for preventing prolonged POI 28 and 5-HT4R agonists that are benzamide analogues of cisapride and MOS also reduce POI, whereas the partial 5-HT4R agonistic and dopamine D2 antagonistic agent metoclopramide does not have preventive action.28 30 These results suggest that the ameliorative actions of cisapride and MOS are not simply mediated by gastrointestinal prokinetic actions through 5-HT4R, and that other mechanisms are involved. In the present study, therefore, we investigated how 5-HT4R agonists prevent POI in a rat model.

MATERIALS AND METHODS Animal model of POI Male SpragueeDawley rats (250e400 g; Charles River Japan, Yokohama, Japan) were manipulated and cared for in strict compliance with the Guide to Animal Use and Care published by the University of Tokyo. The Institutional Review Board of the Graduate School of Agriculture and Life Sciences of the University of Tokyo approved the study protocol. All animals were anaesthetised with pentobarbital sodium (ScheringePlough Corp., Kenilworth, New Jersey, USA) and the animal model of POI was made by intestinal manipulation (IM) previously reported.15e19

Experimental design The rats were randomly assigned to the following groups: Control, no treatment with fasting; IM + Vehicle, sterilised Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

physiological saline was subcutaneously injected at 2 h before and at 2 and 6 h after IM; IM + MOS and IM+CJ, 5-HT4R agonist MOS citrate (0.3, 1 mg/kg, donated by Dainippon Sumitomo Pharma) or CJ-033466 (CJ; 1 mg/kg, donated by Pﬁzer) were similarly administered three times, respectively; IM + MOS + GR, the 5-HT4R antagonist GR113808 (GR; 1 or 3 mg/kg, Sigma Aldrich, St. Louis, Missouri, USA) was similarly administered by three intraperitoneal (i.p.) injections with MOS; IM + MOS + HEX, the non-speciﬁc autonomic ganglionic blocker hexamethonium chloride (1 mg/kg; Nacalai Tesque, Kyoto, Japan) was applied at 1 h before IM and MOS was applied at 2 h before and at 2 and 6 h after IM; IM + MOS + MLA, the a7-nicotinic acetylcholine receptor (a7nAChR) antagonist methyl lycaconitine (MLA) citrate (0.087 mg/kg; TOCRIS, Bristol, UK) was injected i.p. at 30 min before each MOS application. Both MOS and GR113808 were dissolved in 1% of lactic acid with physiological saline and other substances were dissolved in distilled water.

Contraction studies The manipulated ileal portion was isolated from POI model rats at 24 h after IM. The ileum was cut open along the mesenteric attachment, and the mucosal and submucosal layers were gently removed. Circular strips were suspended along the circular axis in a tissue bath ﬁlled with a normal physiological salt solution comprising (in mM) 136.9 NaCl, 5.4 KCl, 1 CaCl2, 1.5 MgCl2, 23.8 NaHCO3, 5.5 glucose, and 0.01 EDTA (pH 7.4). Muscle strips were maintained at 378C and aerated with 95% O2e5% CO2. Responses of the strips were measured isometrically under a resting tension of 10 mN. The magnitude of absolute force was normalised to the wet weight of each strip.

Intestinal transit determination After 18 h of fasting, the rats were randomly assigned to four groups (Control, IM + Vehicle, IM + MOS (MOS 1 mg/kg), IM + MOS + MLA (MOS 1 mg/kg, MLA 0.087 mg/kg) to measure gastrointestinal transit. Twenty-two hours after IM, 200 ml of the non-absorbable marker 0.25% (w/v) phenol red in 5% (w/v) glucose was orally administered to the rats via a gastric tube. After 1 h later, the gastrointestinal part was isolated and stomach and intestine were separated as a single segment (Sto) and divided into ten segments (Sl1eSl10), respectively. The measurement of volume of each segment and calculation of the geometric center of distribution of phenol red were performed as previously reported.17 18

Whole mount immunohistochemistry Whole mount muscularis preparations were basically performed in an orderly manner previously reported.23 24 Each ﬁrst and second antibody are listed in table 1. Preparations were Table 1

Antibodies used for immunohistochemistry

Antibody

Dilution

Anti-ED1mouse monoAb (BMA) Alexa Fluor 488 conjugated anti-mouse IgG donkey polyAb (Invitrogen) Alexa Fluor 568 conjugated anti-mouse IgG donkey polyAb (Invitrogen) Anti-ED2 mouse monoAb (BMA) Alexa Fluor 488 conjugated anti-mouse IgG donkey polyAb (Invitrogen) Anti-PGP9.5 rabbit polyAb (UltraClone) Alexa Fluor 568 conjugated anti-rabbit IgG goat polyAb (Invitrogen) Anti-a7nAChR(C-20) goat polyAb (Santa Cruz Biotechnology) Alexa Fluor 488 conjugated anti-goat IgG rabbit polyAb (Invitrogen) FITC-conjugated a-bungarotoxin (Biotium, Hayward)

1:500 1:500 1:500 1:500 1:500 1:3000 1:500 1:200 1:500 5 mg/ml

639

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology Table 2 sizes

Sequences of PCR primers and their Tm values and product

Gene

Sequences (S: sense, As: antisense)

Tm (8C)

Product size (bp)

GAPDH

S: 59 -TCCCTCAAGATTGTCAGCAA-39 As: 59 -AGATCCACAACGGATACA TT-39 S: 59 -TCCATGAGCTTTGTACAAGG-39 As: 59 -GGTGCTGATGACCAGTTGG-39 S: 59 -CAAGAGACT TCCAGCCAGTTGC-39 As: 59 -TTGCCGAGTAGACCTCATAGTGAC-39 S: 59 -AAATGGGCTCCCTCTCATCA-39 As: 59 -AGCCTTGTCCCTTGAAGAGA-39 S: 59 -CAACTCTCACTGAAGCCAGA-39 As: 59 - AAATGGATCTACATCTTGCA-39 S: 59 -AAGGGAGTGTTGTTCCAGGT-39 As: 59 -CCACCAGCTTCTTCAAAGTG -39 S: 59 -CTTCGGTGCCATTGAGTTG-39 As: 59 -GCATAGGGCTTGTTGACCAT-39 S: 59 -ATGGTGGCAAATGCCTAAG-39 As: 59 -CTCGGAAGCCAATGTAGAGC-39

58

308

52

246

48

614

52

248

54.5

600

52

196

58

354

58

204

IL-1b IL-6 TNFa MCP-1 iNOS 5-HT4R a7-nAChR

immunohistochemically analysed using an LSM510 confocal microscope (Carl Zeiss Japan, Tokyo, Japan) and a Digital Eclipse C1 (Nikon, Tokyo, Japan). Immunopositive cells in the myenteric plexus and the subserosal layers of three randomly selected areas in each preparation were counted and then the averaged value was calculated. The same experiment was performed at least four times to calculate means 6 SE.M.

Myeloperoxidase staining Whole mount preparations of ileal muscularis region were cut into 0.531 cm pieces and ﬁxed in 4% paraformaldehyde in phosphate-buffered saline (PBS) for 30 min at 48C. The preparations were washed with Tris-buffered saline (TBS) for 1 h, incubated in physiological salt solution containing 0.1% (w/v) HankereYates reagent (Polysciences, Warrington, Pennsylvania, USA) and 0.03% (v/v) hydrogen peroxidase (Mitsubishi Gas Chemical Company, Tokyo, Japan) for 10 min and then rinsed

Figure 1 Effects of 5-HT4R agonists on motility dysfunction and leucocyte infiltration induced by intestinal manipulation (IM). (A) Effects of IM on carbachol-induced contractions in intestinal circular smooth muscles and effect of mosapride citrate (MOS) and CJ-033466 (CJ) on contractility. **Significantly different from IM (n¼4 each). (B) Typical results of whole mount immunohistochemical staining of myenteric plexus region to detect ED2-positive resident macrophages, ED1-positive monocyte-derived macrophages and MPO-positive neutrophils. Red signals indicate PGP9.5-positive myenteric nerves. Green signals indicate ED2- or ED1-positive macrophages. (CeE) Quantification of leucocytes in subserosal and myenteric plexus regions; # and ##, significantly different from control at p<0.05 or p<0.01, respectively; * and **, significantly different from number of leucocytes in IM intestine at p<0.05, and p<0.01, respectively. MOS (0.3 or 1 mg/ml, s.c.) and CJ-033466 (1 mg/kg, s.c.) were administered as described in Materials and methods. Bars show means 6 SEM from four independent experiments. 640

Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology in PBS for 10 min. Cells that were obviously myeloperoxidase (MPO)-positive (neutrophils) in the muscularis and subserosal layers were counted under a microscope (Nikon ACT-1C for DXM1200; Nikon, Tokyo, Japan) in ﬁve randomly selected areas of each preparation. Some preparations were also analysed immunohistochemically using ED1 or ED2 after MPO staining.

Quantitative RT-PCR Quantitative RT-PCR proceeded as described.24 The sequences of the primer and expected product size are listed in table 2. Ampliﬁcation proceeded in a PCR Thermal Cycler MP (TaKaRa Biomedicals, Shiga, Japan) using 30e34 cycles (two-cycle intervals), each consisting of 988C for 10 s, 52e588C for 30 s, and 728C for 1 min. The products of each cycle were resolved on 2% agarose gels containing 0.1% ethidium bromide, and the optimal number of PCR cycles for quantiﬁcation was selected. The density of detectable ﬂuorescent bands was measured using NIH Image software (Image J, Ver. 1.38x).

populations were signiﬁcantly inhibited in both regions of ileal tissues isolated from the POI model rat intestine treated with MOS and CJ-033466 (ﬁgure 1D, E). The increased number of ED2-positive muscularis resident macrophages after IM was not changed by 5-HT4R stimulation with MOS and CJ-033466 (ﬁgure 1C). Neither MOS nor CJ-033466 at 1 mg/kg, s.c. affected the macrophage and neutrophil populations in the muscle layer of control rats (n¼4 each; data not shown). We further investigated the effects of GR113808, a 5HT4 receptor selective antagonist, on the MOS-induced anti-inﬂammatory reactions and its ameliorative effect on intestinal motility dysfunction (ﬁgure 2). GR113808 (1, 3, 10 mg/kg, i.p.)

Statistics Results are expressed as means 6 SEM. Data were statistically evaluated using unpaired Student t tests for comparisons between two groups and by one-way analysis of variance (ANOVA) followed by Dunnett’s test for comparisons among three or more groups. Values of p<0.05 were considered statistically signiﬁcant.

RESULTS 5-HT4R stimulation ameliorates motility dysfunction induced by intestinal manipulation in a post-operative ileus model rat intestine Carbachol (CCh) concentration-dependently induced and signiﬁcantly diminished contractions in ileal circular smooth muscle tissues isolated from the small intestines of normal and POI model rats, respectively. The nitric oxide synthase inhibitor, NG-nitro-L-arginine methyl ester (L-NAME, 300 mM; n¼4; IM (CCh 1 mM), 0.045560.0090 mN/mg; IM+L-NAME (CCh; mM), 0.271460.0662 mN/mg; p<0.01) almost completely recovered the decreased contractility, suggesting that the motility dysfunction induced by IM is mediated by NO generated from inducible nitric oxide synthase. Figure 1A shows that the decreased intestinal motility was almost completely recovered in the ileal tissue isolated from POI model rats treated with 1 mg/kg s.c. of either of the speciﬁc 5-HT4R agonists MOS or CJ-033466.

5-HT4 receptor stimulation inhibited inflammation induced by intestinal manipulation We immunohistochemically monitored changes in ED2positive resident macrophages, ED1-positive monocyte-derived macrophages, and MPO-stained neutrophils (ﬁgure 1BeE). Many ED2-positive resident macrophages were detected in the myenteric plexus and subserosal regions of the control rat ileal muscle layer (ﬁgure 1B, C). On the other hand, ED1-positive monocytederived inﬁltrating macrophages and MPO-positive inﬁltrating neutrophils were undetectable in control animals (ﬁgure 1B, D). The ED2-positive resident macrophage population was increased by 25% in the inﬂamed muscle layer of the intestine of the POI model rat compared with that of control rats. On the other hand, many ED1-positive monocyte-derived macrophages and MPOstained neutrophils inﬁltrated into the myenteric plexus region and subserosal region 24 h after the intestinal inﬂammation. Interestingly, monocyte-derived macrophages and neutrophil Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

Figure 2 Effect of the 5-HT4R antagonist GR113808 on MOS-induced leucocyte infiltration and motility dysfunction. Numbers of ED1-positive monocyte-derived macrophages (A) and MPO-stained neutrophils (B) in myenteric plexus region; ##, significantly different from control at p<0.01; **, significantly different from IM at p<0.01; jj, significantly different from IM with MOS (IM+MOS). Each bar shows mean6SEM from three independent experiments. (C) Changes in intestinal muscle contraction induced by CCh. MOS (1 mg/kg) and GR113808 (1, 3 and 10 mg/kg) were administered as described in Materials and methods. * and **, significantly different from IM+MOS (n¼4 each). CCh, carbachol chloride; IM, intestinal manipulation; MOS, mosapride citrate; MPO, myeloperoxidase. 641

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology Figure 3 Effects of 5-HT4R activation on expression of inflammatory mediators in inflamed muscle layer of a post-operative ileus (POI) model rat small intestine. Detailed procedures and predicted sizes of PCR products are described in Materials and methods. ##, significantly different from control at p<0.01; ##, significantly different from control at p<0.01; **, significantly different from intestinal manipulation (IM). Each bar shows mean 6 SEM from four independent experiments.

abolished the inhibitory effect of MOS on the inﬁltration of monocyte-derived macrophages and neutrophils into the inﬂamed muscle layer (ﬁgure 2A, B). In addition, GR113808 suppressed the ameliorative effect of MOS on the IM-mediated motility dysfunction (ﬁgure 2C). GR113808 (1 mg/kg, i.p.) did not affect the populations of ED1-positive macrophages and neutrophils in the muscle layer of control rat intestine (n¼4, data not shown). To evaluate the inﬂammation in the muscle layer of ileum after IM, we investigated changes in interleukin-1 b (IL-1b), IL-6, tumour necrosis factor a (TNF-a), monocyte chemoattractant protein-1 (MCP-1) and inducible nitric oxide synthase (iNOS) levels at 6 h after IM by quantitative RT-PCR. The mRNA expression of IL-1b, MCP-1 and iNOS was signiﬁcantly elevated by IM, which was remarkably attenuated by MOS. The mRNA expression of TNFa and IL-6 tended to increase and MOS also inhibited the tendency (ﬁgure 3).

Anti-inflammatory reaction induced by 5-HT4R stimulation is caused by a neurogenic reaction Stimulating the 5-HT4R activates myenteric plexus cholinergic neurons to release acetyl choline (ACh), which in turn induces gastroprokinetic action in the gastro intestine.4 5 Therefore,

we used the autonomic ganglionic blocker HEX to investigate whether the MOS-mediated anti-inﬂammatory actions are neurogenically mediated (ﬁgures 4 and 5). HEX (1 mg/kg, i.p.) did not affect the populations of ED1-positive macrophages and MPO-positive neutrophils in the myenteric plexus or in the subserosal region of the control rat intestine (n¼4; values are cells/mm2; ED1: myenteric, 6.7666.43; subserosa, 5.4661.22; MPO: myenteric 23.3266.98; subserosa, 5.3961.42). HEX also did not affect the increase in inﬁltration by ED1-positive macrophages and neutrophils induced by IM (n¼4; values are cells/mm2; ED1: myenteric, 1245.926186.57; subserosa, 1150.936115.43; MPO: myenteric, 911.05660.49; subserosa, 1011.19683.68). However, HEX signiﬁcantly reduced the ability of MOS to inhibit inﬁltration by ED1-positive macrophages and MPO-stained neutrophils (ﬁgures 4 and 5A,B). We then investigated the effect of HEX on the MOS-induced amelioration of motility dysfunction by IM (ﬁgure 5C). HEX (1 mg/kg, i.p.) did not affect the CCh (1 mm)-induced contractility of ileal tissue isolated from normal and POI model rats (n¼4 each; values are mN/mg; normal rat, 0.281560.0648; normal rat with HEX, 0.236360.0575; POI model rat, 0.044060.0090; POI model rat with HEX, 0.075060.0156). However, HEX abolished the ability of MOS to improve the motility dysfunction in the POI model rat intestine.

Figure 4 Effects of the autonomic ganglionic blocker hexamethonium and a7nAChR antagonist methyl lycaconitine (MLA) on infiltration of ED1positive monocyte-derived macrophages and myeloperoxidase (MPO)-stained neutrophils into subserosal region after intestinal manipulation (IM). Data are typical findings from four independent experiments. Mosapride (MOS; 1 mg/ kg, s.c.), hexamethonium (1 mg/kg, i.p.) and MLA (0.087 mg/kg, i.p.) were administered as described in Materials and methods.

642

Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology

Figure 5 Quantification of antagonistic effects of hexamethonium and MLA on MOS-induced anti-inflammatory activities determined as leucocyte infiltration and intestinal motility dysfunction in POI model rats. Numbers of ED1-positive monocyte-derived macrophages (A) and MPOstained neutrophils (B). #, ** and jj, significantly different from control, IM (IM), and IM plus MOS (IM + MOS), respectively. Bars indicate means 6 SEM from four independent experiments. Changes in intestinal muscle contraction by CCh (C). MOS citrate (1 mg/kg, s.c.), hexamethonium (1 mg/kg, i.p.) and MLA citrate (0.087 mg/kg, i.p.) were administered as described in Materials and methods; **, significantly different from IM+MOS (n¼4 each). CCh, carbachol chloride; IM, intestinal manipulation; MLA, methyl lycaconitine; MOS, mosapride citrate; MPO, myeloperoxidase; POI, post-operative ileus.

Methyl lycaconitine citrate, an a7nAChR antagonist, abolished the anti-inflammatory action induced by 5-HT4R stimulation Cholinergic neuronal stimulation induces immuno-suppressive actions through a7nAChR on immunoreactive cells such as macrophages and T cells,31 suggesting that 5-HT4R stimulation activates cholinergic neurons to release ACh, which may secondarily result in a7nAChR activation in immunoreactive cells. We therefore investigated the effect of the a7nAChR Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

antagonist methyl lycaconitine citrate (MLA) on MOS-induced anti-inﬂammatory reactions in the POI model rat intestine. We conﬁrmed that MLA (0.087 mg/kg, i.p.) did not affect inﬁltration by ED1-positive macrophages and MPO-stained neutrophils in control and POI model rat intestines (n¼4; data not shown). However, MLA completely suppressed the MOS-induced anti-inﬂammatory activity determined as macrophage and neutrophil inﬁltration into the muscle layer (ﬁgures 4 and 5A, B). We also examined effect of MLA on the MOS-induced ameliorative effect on IM-mediated motility dysfunction (ﬁgure 5C). MLA (0.087 mg/kg i.p.) did not affect the CCh (1 mM)-induced contraction of ileal tissue in control and POI model rats (n¼4 each; values are mN/mg; normal rat, 0.281560.0648; normal rat with MLA, 0.251260.0291; postoperative ileus model rat, 0.044060.0090; postoperative ileus model rat with MLA, 0.056560.0101). Figure 5C shows that MLA (0.087 mg/kg i.p.) inhibited the ameliorative action of MOS on intestinal dysmotility caused by IM. To conﬁrm the effect of MLA on MOS-induced ameliorative action for IM-mediated gastrointestinal motility disorder in vivo, gastrointestinal transit was measured at 22e23 h after IM. About 30% of the orally administered phenol red labelled content remained inside the stomach and 70% of it was transported down the intestine to the distal end of ileum with a peak (SI9) in control healthy rats (ﬁgure 6A). The averaged calculated geometric centre for the control animals was 6.6360.41 for 11 segments of the gastrointestinal tract (ﬁgure 6E). In contrast, IM caused a signiﬁcant delay in gastrointestinal transit after a 1-h transit period, and 70% of the orally administered content remained in the stomach, whereas 30% of the transported content was moved around the jejunum and proximal ileum. The geometric centre was 2.6060.49 (n¼4, ﬁgure 6B, E). The administration of MOS signiﬁcantly recovered the delayed gastrointestinal transit and reduced the value of the geometric centre after IM (ﬁgure 6C, E). Furthermore, MLA (0.087 mg/kg s.c.) signiﬁcantly inhibited the ameliorative action of MOS on the delayed gastrointestinal transit caused by IM (ﬁgure 6D, E) in which 50% of the orally administered content remained in the stomach, and 50% of the transported content was moved around the jejunum and proximal ileum (geometric centre value: 3.4760.61, n¼4). MOS did not affect gastrointestinal transit of control healthy rat (6.7360.92, n¼4), suggesting that MOS does not induce gastrointestinal prokinetic action in the current experimental condition.

ED1- and ED2-positive macrophages express a7nAChR whereas neutrophils do not We investigated which cells in the intestinal wall express a7nAChR cells after IM using a-bungarotoxin (a-BTX) conjugated with ﬂuorescein isothiocyanate (FITC) (ﬁgures 7 and 8). We rarely detected a-BTX-bound cells in the myenteric plexus and serosal regions of control ileal tissues (myenteric plexus and subserosal regions, 10.461.86 and 5.6661.72 cells/mm2, respectively; n¼4). In contrast, many a-BTX-positive cells were detected in both regions of inﬂamed ileal tissues (myenteric plexus and subserosal regions, 760.5640.67 and 750.49659.53 cells/mm2; n¼4 each). We double-stained specimens of the inﬂamed muscle layer with ED1- and ED2-antibody or MPO and FITC-labelled a-BTX. Over 50% of the population of round ED2-positive activated resident macrophages bound to a-BTX and the ratios of ED1-positive inﬁltrating macrophages that bound to a-BTX were similar in both regions of the inﬂamed muscle layer (ﬁgures 8). In contrast, MPO-positive neutrophils did not react with a-BTX (ﬁgure 7). 643

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology Figure 6 Ameliorative action of 5-HT4R activation and negative effect of MLA on gastrointestinal transit in POI model rats. (AeD) Detailed procedures are described in Materials and methods. Gastrointestinal transit 1 h after food intake was measured. Each bar indicates means 6 SEM from four independent experiments. (E) Values of geometric centre of four groups (control, IM + vehicle, IM + MOS, IM + MOS + MLA) were shown. ##, ** and jj, significantly different from control, IM (IM), IM plus MOS (IM + MOS), respectively. IM, intestinal manipulation; MLA, methyl lycaconitine; MOS, mosapride citrate; POI, post-operative ileus.

We further performed immunohistochemical double staining using anti-a7nAChR and anti-ED1 or anti-ED2 antibodies. Many cells were double stained with anti-a7nAChR and antiED1 or anti-ED2 antibodies in the inﬂamed myenteric plexus region at 24 h after IM (ﬁgure 9A, B: ED2 and anti-a7nAChR, 136.61627.98 cells per mm2; ED1 and anti-a7nAChR, 256.44648.21 cells per mm2. n¼4 each). We also performed double staining using anti-a7nAChR antibody and a-BTX. The results indicated that almost all anti-a7nAChR antibodypositive and a-BTX-positive cells merged (ﬁgure 9C, n¼4).

DISCUSSION Muscularis inﬂammation induces a motility disorder at the prolonged phase of POI.12 We demonstrated here that the gastroprokinetic agent MOS ameliorates the motility dysfunction in the POI. Furthermore, we showed that MOS signiﬁcantly suppressed macrophage and neutrophil inﬁltration into the inﬂamed region, suggesting that an anti-inﬂammatory reaction is involved. It is possible that the ameliorative action of MOS on the motility dysfunction might be induced by gastrointestinal prokinetic ability due to MOS remaining in the isolated ileal tissue under assay conditions. However, this is unlikely because rats rapidly eliminate MOS with a t½ of 1.9 h,32 and we isolated intestinal tissues at 18 h after the ﬁnal administration of MOS (at the time of measuring contractile activity), when the MOS concentration in the muscle tissue would be insufﬁcient to induce a prokinetic reaction. We then questioned whether the ameliorative actions of MOS both on motility dysfunction and 644

inﬂammation are mediated by selective action through the 5-HT4R, because MOS metabolites have antagonistic effects on 5-HT3 receptors.33 The potent and selective 5-HT4R agonist CJ-03346634 35 ameliorated the motility dysfunction and the inﬁltration of leucocytes induced by IM. In addition, the selective 5-HT4R antagonist GR113808 36 abolished the effects of MOS. We thus concluded that 5-HT4R stimulation restores the motility dysfunction in POI via an anti-inﬂammatory mechanism that is independent of prokinetic ability. Immune reactive cells such as dendritic cells also express 5-HT4R.37 Activation of the 5-HT4R inhibits TNFa but increases the production of IL-1b and IFNg. Thus, 5-HT4R agonists might directly act on inﬂammatory cells such as macrophages and neutrophils. Therefore, we next examined whether the effect of 5-HT4R agonists is mediated through direct actions on these immune cells or through the neurogenic mechanisms. We found that the non-speciﬁc autonomic ganglionic blocker HEX completely suppressed the 5-HT4R stimulation-mediated antiinﬂammatory reaction, suggesting that 5-HT4R stimulation in POI exerts neuronal anti-inﬂammatory actions. Unlike the observation in human dendritic cells,37 5-HT4R mRNA expression was undetectable in rat peritoneal macrophages (supplementary ﬁgure 1). Regarding the neurogenic mechanism of anti-inﬂammatory actions induced by 5-HT4R agonists, recent understanding of the control mechanisms of intestinal inﬂammation exerted by autonomic nervous systems should be considered. For instance, Tanila and Kauppila reported that a selective a2-adrenergic antagonist reversed laparotomy-induced Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology

Figure 8 Summary of cell populations stained with a-bungarotoxin and ED1- or ED2-positive macrophages in myenteric plexus and subserosal region isolated from normal and POI model rat ileum. Each column shows mean 6 SEM from four independent experiments. The quantitative method is shown in the Materials and methods.

Figure 7 Double staining with a-bungarotoxin and ED1- or ED2positive macrophages or myeloperoxidase (MPO)-stained neutrophils in myenteric plexus region isolated from normal and a post-operative ileus (POI) model rat ileum. Data are typical findings from four independent experiments. Green signals, a-bungarotoxin-bound cells possibly indicating a7nAChR expression; red signals, anti-ED1 or anti-ED2 antibody stained macrophages. ileus, even at the prolonged phase of POI.38 Kreiss et al. 39 demonstrated that macrophages inﬁltrating the muscularis after IM express a2-adrenergic receptors. Furthermore, RAW264.7 macrophages are capable of synthesising catecholamines,40 suggesting that released catecholamines can react with a2-adrenergic receptors on macrophages in an autocrine and a paracrine manner to complicate POI. As an alternative autonomic regulation of inﬂammation, several investigators have suggested that vagus nerve stimulation attenuates gastrointestinal inﬂammations.41e45 Wang et al recently found that nicotine inhibits the production of pro-inﬂammatory cytokines from macrophages more efﬁciently than ACh.42 On the other hand, Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

many type of immune cells such as lymphocytes, dendritic cells and monocyte/macrophages are now recognised to express a7nAChR.31 The activation of a7nAChR triggers a spectrum of anti-inﬂammatory signalling mechanisms that directly or indirectly inhibit NF-kB activation and/or Jak/STAT3 pathway in inﬂammatory cells.31 We found here that the a7nAChR antagonist MLA almost completely suppressed the anti-inﬂammatory action mediated by 5-HT4R stimulation. The amelioration of intestinal dysmotility by 5-HT4R agonists was also abolished by MLA. These results suggest that a7nAChR is involved in the ameliorative action of 5-HT4R stimulation on POI. The present study focused on the role of monocyte/macrophage lineage cells because evidence suggests that these cells express abundant a7nAChR.31 Our immunohistochemical study of inﬂamed intestinal tissues showed that some ED1- and ED2positive macrophages, but not MPO-stained neutrophils, had binding afﬁnity for a-BTX and were stained with anti-a7nAChR antibody. Although the variety of nAChR subunits allows for a diversity of combinations, the MLA-sensitive receptors with high afﬁnity for a-BTX could be a7nAChR, because a-BTX or MLA has binding afﬁnity for the a1, a7 and a9, or a6 and a7 isoforms of nAChR.46 We further analysed the distribution of a7nAChR in more detail in control and inﬂamed small intestine musculature. We detected only a small population of a-BTX or anti-a7nAChR antibody-positive leucocytes in the myenteric plexus and subserosal regions of the normal rat small intestine. 645

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology

Figure 10 Anti-inflammatory mechanisms of 5-HT4R stimulation in post-operative ileus (POI). Stimulation of 5-HT4R in myenteric plexus nerve results in release of acetyl choline (ACh), which in turn activates a7nAChR on activated resident macrophages and infiltrating monocytederived macrophages (but not on neutrophils), which prevents leucocyte infiltration and subsequently inhibits inflammation and motility disorders. ACh released from myenteric plexus nerve might also act on a7, a9 and a10 isoforms of nAChRs of mast cells to prevent inflammation (see de Jonge et al55 and Mishra et al56). Figure 9 Double staining with anti-a7nAChR antibody and ED1- or ED2-positive macrophages in the myenteric plexus region isolated from normal and post-operative ileus (POI) model rat ileum. Data are typical findings from five independent experiments. (A and B) Green signals, anti-ED2 or anti-ED1 antibody-positive macrophages; red signals, antia7nAChR antibody-positive cells. (C) Immunohistochemistry stained with anti-a7nAChR antibody and a-bungarotoxin (BTX). Green signals, FITC conjugated a-bungarotoxin-positive cells; red signals, antia7nAChR antibody-positive cells. In contrast, 24 h after the inﬂammation induced by IM, some ED2-positive activated resident macrophages with a round form and ED1-positive monocyte-derived inﬁltrating macrophages expressed a7nAChR in POI model rats. ED2-positive activated resident macrophages are derived from self-multiplication at the early stage of intestinal inﬂammation.47 ED1-positive inﬁltrating macrophages might also transform to ED2-positive resident macrophages, which become stainable for both ED1 and ED2.47e49 These reports also indicate the possibility that ED1 and ED2 double positive macrophages may also express a7nAChR. Although many reports indicate that a7nAChR is expressed in macrophages and neutrophils, several others suggest otherwise. Therefore, Van Der Zanden and colleagues posited that a7nAChR expression on leucocytes is questionable.50 We did not detect a7nAChR mRNA expression in rat resident peritoneal macrophages (supplementary ﬁgure 1). In contrast, a7nAChR mRNA was detectable in peritoneal macrophages stimulated with LPS (1 mg/ml) for 20 h, although not at very high levels. Recent studies have also indicated that a7nAChR expression is upregulated in macrophage-like U937 cells stimulated with LPS51 and in alveolar macrophages by acid-induced acute lung injury.52 Taken together, we support the idea that a7nAChR expression is modulated by the maturation and transformation of either resident or monocyte-derived intestinal muscularis macrophages by inﬂammatory mediators. 646

Neutrophils apparently express nicotinic receptors, although whether one of these receptor types is a7nAChR remains unclear.31 53 One recent study found that neutrophils in the injured lung express a7nAChR.52 However, neutrophils expressing a7nAChR are unlikely in the inﬂamed intestine, because a-BTX did not bind to inﬁltrated neutrophils stained with MPO in the present study. On the other hand, Saeed et al reported that microvascular endothelial cells express a7nAChR.54 They clariﬁed that vagus nerve stimulation and cholinergic agonists signiﬁcantly block leucocyte migration through a7nAChR in the carrageenan air pouch model. However, we did not detect microvascular tubes that were both a-BTX-positive and stained with anti-a7nAChR antibody in the myenteric plexus region of the small intestine either before or after inducing inﬂammation by IM. We found in this study that 5-HT4R stimulation of myenteric neurons ameliorates intestinal inﬂammation induced by IM, which results in recovered motility function. The 5-HT4R might mediate anti-inﬂammatory actions by stimulating cholinergic neurons of the myenteric plexus to release ACh, which in turn activates a7nAChR on muscularis resident macrophages and monocyte-derived inﬁltrating macrophages to suppress inﬂammation due to IM (ﬁgure 10). On the other hand, it has been suggested that intestinal muscularis mast cells play a pivotal role in the inﬂammation induced by IM55; that is, mast cell activation induces leucocyte inﬁltration to accelerate muscularis inﬂammation in POI. A recent report has shown that the RBL2H3 rat mast cell line expresses a7, a9 and a10 isoforms of nAChR.56 The authors suggested that these three isoforms functionally interact, indicating the possibility of a hybrid nAChR. Therefore, the released ACh might also act on the nAChRs of mast cells to inhibit leucocyte inﬁltration. Further investigation is necessary to clarify this point. In conclusion, although 5-HT4R agonists such as MOS are clinically validated as therapies for gastrointestinal disorders Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Neurogastroenterology characterised by dysmotility, the present ﬁndings provide new insights indicating that 5-HT4R agonists can also serve as anti-inﬂammatory agents to treat diseases associated with gastrointestinal motility.

26.

Acknowledgements We thank Dainippon Sumitomo Pharma Co. Ltd. and Pfizer Inc. for supplying MOS and CJ-033466, respectively.

29.

Funding This work was supported by Grants-in-Aid for Scientific Research from the Japanese Ministry of Education (to MH, no.18380173 and no. 21380178; to HO, no. 20228005; and to TM, no. 19688014).

30.

Competing interests HO received grant support from Dainippon Sumitomo Pharma Co. Ltd. The remaining authors have declared no financial interests.

27. 28.

31. 32.

Patient consent Not needed. Provenance and peer review Not commissioned; externally peer reviewed.

33.

REFERENCES

34.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

Abell TL, Camilleri M, DiMagno EP, et al. Long-term efficacy of oral cisapride in symptomatic upper gut dysmotility. Dig Dis Sci 1991;36:616e20. Appel-Dingemanse S. Clinical pharmacokinetics of tegaserod, a serotonin 5-HT4 receptor partial agonist with promotile activity. Clin Pharmacokinet 2002;41:1021e42. Deruyttere M, Lepoutre L, Heylen H, et al. Cisapride in the management of chronic functional dyspepsia: a multicenter, double-blind, placebo-controlled study. Clin Ther 1987;10:44e51. Kim HS, Choi EJ, Park H. The effect of mosapride citrate on proximal and distal colonic motor function in the guinea-pig in vitro. Neurogastroenterol Motil 2008;20:169e76. Cellek S, John AK, Thangiah R, et al. 5-HT4 receptor agonists enhance both cholinergic and nitrergic activities in human isolated colon circular muscle. Neurogastroenterol Motil 2006;18:853e61. Baker DE. Rationale for using serotonergic agents to treat irritable bowel syndrome. Am J Health Syst Pharm 2005;62:700e11; quiz 12e13. Gershon MD, Tack J. The serotonin signaling system: from basic understanding to drug development for functional GI disorders. Gastroenterology 2007;132:397e414. Liu M, Geddis MS, Wen Y, et al. Expression and function of 5-HT4 receptors in the mouse enteric nervous system. Am J Physiol Gastrointest Liver Physiol 2005;289: G1148e63. Poole DP, Xu B, Koh SL, et al. Identification of neurons that express 5hydroxytryptamine 4 receptors in intestine. Cell Tissue Res 2006;325:413e22. Prins NH, Akkermans LM, Lefebvre RA, et al. 5-HT4 receptors on cholinergic nerves involved in contractility of canine and human large intestine longitudinal muscle. Br J Pharmacol 2000;131:927e32. Livingston EH, Passaro EP Jr. Postoperative ileus. Dig Dis Sci 1990;35:121e32. Bauer AJ, Boeckxstaens GE. Mechanisms of postoperative ileus. Neurogastroenterol Motil 2004;16(Suppl 2):54e60. Boeckxstaens GE, Pelckmans PA, Rampart M, et al. Pharmacological characterization of 5-hydroxytryptamine receptors in the canine terminal ileum and ileocolonic junction. J Pharmacol Exp Ther 1990;254:652e8. De Winter BY, Boeckxstaens GE, De Man JG, et al. Effect of adrenergic and nitrergic blockade on experimental ileus in rats. Br J Pharmacol 1997;120:464e8. Kalff JC, Carlos TM, Schraut WH, et al. Surgically induced leukocytic infiltrates within the rat intestinal muscularis mediate postoperative ileus. Gastroenterology 1999;117:378e87. Kalff JC, Schraut WH, Billiar TR, et al. Role of inducible nitric oxide synthase in postoperative intestinal smooth muscle dysfunction in rodents. Gastroenterology 2000;118:316e27. Schwarz NT, Kalff JC, Turler A, et al. Prostanoid production via COX-2 as a causative mechanism of rodent postoperative ileus. Gastroenterology 2001;121:1354e71. Wehner S, Behrendt FF, Lyutenski BN, et al. Inhibition of macrophage function prevents intestinal inflammation and postoperative ileus in rodents. Gut 2007;56:176e85. Schmidt J, Stoffels B, Moore BA, et al. Proinflammatory role of leukocyte-derived Egr-1 in the development of murine postoperative ileus. Gastroenterology 2008;135:926e36 e1e2. Mikkelsen HB, Thuneberg L, Rumessen JJ, et al. Macrophage-like cells in the muscularis externa of mouse small intestine. Anat Rec 1985;213:77e86. Ozaki H, Kawai T, Shuttleworth CW, et al. Isolation and characterization of resident macrophages from the smooth muscle layers of murine small intestine. Neurogastroenterol Motil 2004;16:39e51. Mikkelsen HB, Rumessen JJ. Characterization of macrophage-like cells in the external layers of human small and large intestine. Cell Tissue Res 1992;270:273e9. Kinoshita K, Horiguchi K, Fujisawa M, et al. Possible involvement of muscularis resident macrophages in impairment of interstitial cells of Cajal and myenteric nerve systems in rat models of TNBS-induced colitis. Histochem Cell Biol 2007;127:41e53. Hori M, Kita M, Torihashi S, et al. Upregulation of iNOS by COX-2 in muscularis resident macrophage of rat intestine stimulated with LPS. Am J Physiol Gastrointest Liver Physiol 2001;280:G930e8. Holte K, Kehlet H. Postoperative ileus: a preventable event. Br J Surg 2000;87:1480e93.

Gut 2011;60:638e647. doi:10.1136/gut.2010.227546

35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

Seta ML, Kale-Pradhan PB. Efficacy of metoclopramide in postoperative ileus after exploratory laparotomy. Pharmacotherapy 2001;21:1181e6. Sanger GJ. 5-Hydroxytryptamine and the gastrointestinal tract: where next? Trends Pharmacol Sci 2008;29:465e71. Zeinali F, Stulberg JJ, Delaney CP. Pharmacological management of postoperative ileus. Can J Surg 2009;52:153e7. De Winter BY, Boeckxstaens GE, De Man JG, et al. Effect of different prokinetic agents and a novel enterokinetic agent on postoperative ileus in rats. Gut 1999;45:713e18. Narita K, Tsunoda A, Takenaka K, et al. Effect of mosapride on recovery of intestinal motility after hand-assisted laparoscopic colectomy for carcinoma. Dis Colon Rectum 2008;51:1692e5. de Jonge WJ, Ulloa L. The a7 nicotinic acetylcholine receptor as a pharmacological target for inflammation. Br J Pharmacol 2007;151:915e29. Sakashita M, Mizuki Y, Hashizume T, et al. Pharmacokinetics of the gastrokinetic agent mosapride citrate after intravenous and oral administrations in rats. Arzneimittelforschung 1993;43:859e63. Yoshida N, Omoya H, Kato S, et al. Pharmacological effects of the new gastroprokinetic agent mosapride citrate and its metabolites in experimental animals. Arzneimittelforschung 1993;43:1078e83. Mikami T, Ochi Y, Suzuki K, et al. 5-Amino-6-chloro-N-[(1-isobutylpiperidin-4-yl) methyl]-2-methylimidazo[1,2-a]pyridine-8-carboxamide (CJ-033,466), a novel and selective 5-hydroxytryptamine 4 receptor partial agonist: pharmacological profile in vitro and gastroprokinetic effect in conscious dogs. J Pharmacol Exp Ther 2008;325:190e9. Toga T, Kohmura Y, Kawatsu R. The 5-HT4 agonists cisapride, mosapride, and CJ-033466, a novel potent compound, exhibit different human ether-a-go-go-related gene (hERG)-blocking activities. J Pharmacol Sci 2007;105:207e10. Grossman CJ, Kilpatrick GJ, Bunce KT. Development of a radioligand binding assay for 5-HT4 receptors in guinea-pig and rat brain. Br J Pharmacol 1993;109:618e24. Idzko M, Panther E, Stratz C, et al. The serotoninergic receptors of human dendritic cells: identification and coupling to cytokine release. J Immunol 2004;172:6011e19. Tanila H, Kauppila T, Taira T. Inhibition of intestinal motility and reversal of postlaparotomy ileus by selective a2-adrenergic drugs in the rat. Gastroenterology 1993;104:819e24. Kreiss C, Toegel S, Bauer AJ. a2-Adrenergic regulation of NO production alters postoperative intestinal smooth muscle dysfunction in rodents. Am J Physiol Gastrointest Liver Physiol 2004;287:G658e66. Brown SW, Meyers RT, Brennan KM, et al. Catecholamines in a macrophage cell line. J Neuroimmunol 2003;135:47e55. Borovikova LV, Ivanova S, Zhang M, et al. Vagus nerve stimulation attenuates the systemic inflammatory response to endotoxin. Nature 2000;405:458e62. Wang H, Yu M, Ochani M, et al. Nicotinic acetylcholine receptor a7 subunit is an essential regulator of inflammation. Nature 2003;421:384e8. Ghia JE, Blennerhassett P, Kumar-Ondiveeran H, et al. The vagus nerve: a tonic inhibitory influence associated with inflammatory bowel disease in a murine model. Gastroenterology 2006;131:1122e30. The FO, Boeckxstaens GE, Snoek SA, et al. Activation of the cholinergic anti-inflammatory pathway ameliorates postoperative ileus in mice. Gastroenterology 2007;133:1219e28. de Jonge WJ, van der Zanden EP, The FO, et al. Stimulation of the vagus nerve attenuates macrophage activation by activating the Jak2-STAT3 signaling pathway. Nat Immunol 2005;6:844e51. Whiteaker P, Marks MJ, Christensen S, et al. Synthesis and characterization of 125 I-a-conotoxin ArIB[V11L;V16A], a selective a7 nicotinic acetylcholine receptor antagonist. J Pharmacol Exp Ther 2008;325:911e19. Hori M, Nobe H, Horiguchi K, et al. MCP-1 targeting inhibits muscularis macrophage recruitment and intestinal smooth muscle dysfunction in colonic inflammation. Am J Physiol Cell Physiol 2008;294:C391e401. Schwarz NT, Beer-Stolz D, Simmons RL, et al. Pathogenesis of paralytic ileus: intestinal manipulation opens a transient pathway between the intestinal lumen and the leukocytic infiltrate of the jejunal muscularis. Ann Surg 2002;235:31e40. Gordon S, Taylor PR. Monocyte and macrophage heterogeneity. Nat Rev Immunol 2005;5:953e64. Van Der Zanden EP, Boeckxstaens GE, de Jonge WJ. The vagus nerve as a modulator of intestinal inflammation. Neurogastroenterol Motil 2009;21:6e17. Chernyavsky AI, Arredondo J, Skok M, et al. Auto/paracrine control of inflammatory cytokines by acetylcholine in macrophage-like U937 cells through nicotinic receptors. Int Immunopharmacol 2010;10:308e15. Su X, Lee JW, Matthay ZA, et al. Activation of the a7 nAChR reduces acid-induced acute lung injury in mice and rats. Am J Respir Cell Mol Biol 2007;37:186e92. Davies BD, Hoss W, Lin JP, et al. Evidence for a noncholinergic nicotine receptor on human phagocytic leukocytes. Mol Cell Biochem 1982;44:23e31. Saeed RW, Varma S, Peng-Nemeroff T, et al. Cholinergic stimulation blocks endothelial cell activation and leukocyte recruitment during inflammation. J Exp Med 2005;201:1113e23. de Jonge WJ, The FO, van der Coelen D, et al. Mast cell degranulation during abdominal surgery initiates postoperative ileus in mice. Gastroenterology 2004;127:535e45. Mishra NC, Rir-sima-ah J, Boyd RT, et al. Nicotine inhibits FcvarepsilonRI-induced cysteinyl leukotrienes and cytokine production without affecting mast cell degranulation through a7/a9/a10-nicotinic receptors. J Immunol 2010;185:588e96.

647

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon

Flagellin administration protects gut mucosal tissue from irradiation-induced apoptosis via MKP-7 activity Rheinallt M Jones, Valerie M Sloane, Huixia Wu, Liping Luo, Amrita Kumar, Matam Vijay Kumar, Andrew T Gewirtz, Andrew S Neish < Additional figures are

published online only. To view these files please visit the journal online (http://gut.bmj. com). Epithelial Pathobiology Unit, Department of Pathology and Laboratory Medicine, Emory University School of Medicine, Atlanta, Georgia, USA Correspondence to Andrew S Neish, Department of Pathology, Emory University School of Medicine, Room 105F, Whitehead Bldg, 615 Michael Street, Atlanta, GA 30322, USA; [email protected] Revised 30 October 2010 Accepted 4 November 2010 Published Online First 3 January 2011

ABSTRACT Background and aims Radiotherapy for neoplastic disease is associated with significant adverse enteric effects associated with excessive cell death. Ionising radiation induces cell death by a mechanism that is dependent on JNK (c-jun N-terminal kinase) pathway signalling. Additionally, it is known that cells exposed to extracellular bacterial products such as flagellin, pleiotropically activate a number of innate immune pathways, including that of JNK. The JNK pathway controls its own activity by inducing the transcription of mitogen-activated protein kinase phosphatase-7 (MKP-7) which directly targets phosphorylated JNK, thus functioning as a negative feedback loop. Previously, it has been shown that flagellin limits ionising radiationinduced mortality in mice, but the cellular mechanism of protection remained unknown. Methods Wild-type C57BL/6 or tlr5/ C57BL/6 were injected with flagellin 2 h before exposure to irradiation, and their intestines were examined for apoptosis. Candidate proteins mediating cytoprotection from irradiation were identified by expression profiling. One of these candidates, MKP-7, was cloned and packaged into adenovirus particles, used to infect cultured cells, and examined for the extent to which its activity reduced cellular apoptosis by flow cytometry or immunoblot analysis. Results Flagellin pretreatment protected mice from radiation-induced intestinal mucosal injury and apoptosis via a Toll-like receptor 5 (TLR5)-dependent mechanism. Expression profiling of flagellin-treated mice showed upregulation of MKP-7, an inducible repressor of the JNK pathway. MKP-7 expression reached a maximum at 2 h after flagellin treatment, coinciding with suppression of phosphorylated JNK and JNK pathway inhibition. Furthermore, constitutive MKP-7 expression protected cultured cells from radiation-induced apoptosis. Conclusions Flagellin is a promising adjuvant for suppressing ionising radiation-induced injury. MKP-7 activity exhibits cytoprotective effects, and is thus a candidate cellular molecule for limiting the damaging effect of radiotherapy on the gastreointestinal system.

INTRODUCTION Ionising radiation is used as adjuvant or curative treatment in cancer patients to control malignant cell growth. Despite clinical efﬁcacy, irradiation treatment also results in harmful side effects in radiosensitive tissues and cell types with high rates of cell division, including, but not limited to, lymphoid organs, bone marrow, testis, ovaries and the intestine. In the intestine, radiation is known to induce 648

Significance of this study What is already known about this subject?

< Cancer radiotherapy results in harmful side

effects secondary to gut apoptosis.

< Flagellin preadministration to mice has been

reported to have protective effects against irradiation-induced mortality. < Irradiation-induced apoptosis in known to be JNK pathway dependent.

What are the new findings?

< Here, we found that flagellin pre-treatment

protects the intestinal mucosa from injury resulting from irradiation-induced apoptosis. < Flagellin treatment induced transcription of mkp-7 in the intestinal mucosa to levels that reached a maximum at 2 h after flagellin treatment. < Flagellin is most effective at protecting against irradiation-induced apoptosis when administered 2 h pre-exposure, coinciding with the period when levels of MKP-7 are highest. < Constitutive expression of MKP-7 is a potent inhibitor of JNK pathway signalling and irradiation-induced apoptosis.

How might it impact on clinical practice in the foreseeable future? < Our results suggest the specialised use of flagellin as potential adjuvant therapy for limiting irradiation-induced intestinal injury. < MKP-7 is a candidate molecule for the modulation of irradiation-induced apoptosis.

programmed cell death in the mucosa, particularly in proliferating cells at the base of the villous crypts. Within days of exposure, damage to the epithelial mucosa culminates in gastrointestinal syndrome, in which extensive infection and electrolyte imbalance may cause signiﬁcant morbidity and ultimately result in death.1e3 One aim of modern radiotherapy is to reduce the harmful side effects to a minimum. Recent reports have shown that bacterial products and/or proteins as well as cytokines decrease the sensitivity of cells to irradiation-induced apoptosis.4e6 Previous reports have described the phenomenon that pretreatment of mice with the Toll-like receptor 5 (TLR5) agonist ﬂagellin before insult by sublethal doses of ionising radiation results in improved murine survival rates.7e9 However, the molecular mechanism of ﬂagellin (or any bacterially derived compound used as a therapeutic) - mediated Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon protection against ionising radiation injury is unknown, and the pathology of radiosensitive tissues in experimental subjects remain unexamined. Exposure of cells to ionising radiation initiates the activation of cellular signalling pathways that result in apoptosis. The cellular mechanisms of ionising radiation-induced apoptosis have recently been characterised and show an essential role for c-jun N-terminal kinase (JNK) in activating the intrinsic/mitochondrial apoptotic pathway.10e13 Ionising radiation-induced JNK phosphorylation activates the proapoptotic members of the Bcl-2 family of intracellular proteins, including Bax and Bak, which cause mitochondrial dysfunction resulting in the release of proapoptotic factors (eg, cytochrome c) that activate the initiator caspase-9 and terminate in apoptosis.11 13 14 Upregulation of the JNK signalling pathway can be also initiated by a variety of extracellular signals including bacterial products and proteins such as ﬂagellin.15 Similar to other mitogen-activated protein (MAP) kinases, JNK is activated by a phosphorylation cascade module comprised of a MAP kinase kinase and a MAP kinase kinase kinase. Once activated, JNK phosphorylates a number of transcription factors which results in the increased expression of a battery of innate immune- and apoptosis-related genes.14 JNK pathway activity is negatively regulated by dual speciﬁcity phosphatases (DUSPs) which are transcriptionaly activated in response to JNK activation and serve as a negative feedback loop to return JNK signalling to basal levels.16 One member of this family, MAP kinase phosphatase-7 (MKP-7), also known as DUSP16, has activity and high speciﬁcity against phosphorylated JNK.17e20 In this study, we show that systemic (intraperitoneal) administration of puriﬁed ﬂagellin 2 h prior to radiation insult signiﬁcantly reduced the number of apoptotic cells in the gut mucosa. Importantly, we also show that ﬂagellin pretreatment mediates protection against ionising radiation via the transient upregulation of MKP-7 in the cytoplasm, which dephosphorylates JNK, thus inhibiting additional JNK pathway activation that would ensue following ionising radiation insult. Indeed, transfection of cells with small interfering RNA (siRNA) against MKP-7 resulted in elevated levels of phosphorylated JNK and increased apoptosis even in the absence of an external stimulus. Furthermore, constitutive cellular expression of MKP-7 within cells signiﬁcantly decreased irradiation-induced apoptosis. Together, these data show that the mechanism of ﬂagellinelicited cytoprotection against irradiation-induced injury to the gut mucosa is by MKP-7 inhibition of JNK pathway activity. This investigation highlights the cytoprotective activity of MKP-7 and also suggests its potential suitability as a therapeutic molecule.

MATERIALS AND METHODS Murine subjects and g-irradiation All experiments were done using 12-week-old C57BL/6 background mice that had been purchased from The Jackson Laboratory (Bar Harbor, Maine, USA), except for TLR5/ mice, which were generated/maintained as previously described.21 Mice whole bodies were exposed to 8 Gy of g-radiation using a g-Cell 40 137Cs irradiator at a dose rate of 75 rads/min. Flagellin (50 mg) (or the indicated concentration) was administered intraperitoneally 2 h (or the indicated time) preceding (or following) irradiation, and body weights and mortality were monitored. Animal experiments were approved by the Emory University institutional ethical committee and performed according to the legal requirements. Cultured mammalian cells were exposed to 12 Gy of g-radiation using a g-Cell 40 137Cs Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

irradiator at a dose rate of 75 rads/min. Histological sections of small intestine were prepared from three irradiated animals per treatment. Sections were assessed by TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay (Roche, Indianapolis, Indiana, USA) and by immunohistochemistry using antibodies against cleaved caspase-3. TUNELor active caspase-3-positive cells were counted and the average number of positive cells in forty 2003 ﬁelds per treated animal was determined.

Microarray analysis Details of cDNA microarray fabrication, hybridisation, scanning and labelling of RNA samples have been described previously.15 Total RNA from murine subjects was prepared using TRIzol reagent (Invitrogen, Carlsbad, California, USA). With the use of an oligo(dT) primer and Superscript II Reverse Transcriptase (Invitrogen), labelled cDNA was synthesised from 40 mg of total RNA. Reference and experimental RNA samples were labelled with Cy3- and Cy5-coupled dCTP (Amersham Biosciences; Piscataway, New Jersey, USA), respectively, and hybridised to microarrays. Three repeats for each experimental condition were performed. Following normalisation, genes with intensity >4 times the background mean were selected. To minimise variability, the mean of three independent experiments for each gene was calculated and used for ﬁnal data clustering. Only genes that showed signiﬁcant change (over twofold difference) were selected for report. Cluster/Tree view (Michael Eisen, Stanford University, Stanford, California, USA) analytic software packages were used for hierarchical clustering.

RESULTS Flagellin pretreatment reduces irradiation-induced apoptosis in murine intestinal mucosa by a TLR5-dependent mechanism Radiation insult is known to induce premature apoptosis in the mammalian intestinal mucosa, particularly in proliferating cells in villous crypts. Previous reports by our research group and others showed that intraperitoneal administration of ﬂagellin in mice mediates protection from radiation-induced mortality.9 In order to examine the extent to which intraperitoneal ﬂagellin administration prior to radiation insult has protective activity against injury to the intestinal mucosa, we injected ﬂagellin puriﬁed from Salmonella typhimurium SL3201 into wild-type C57BL/6 mice and TLR5/ C57BL/6 mice at 2 h before wholebody insult with 8 Gy of irradiation. Animals were sacriﬁced at 6 h postirradiation and transverse sections of ileal tissue examined for crypt apoptosis by morphological analysis. Irradiated C57BL/6 mice pretreated with phosphate-buffered saline (PBS) exhibited increased numbers of pycnotic cells (condensed nuclei) and karyorrhexis (fragmented nuclei) within the crypt base compared with untreated mice (ﬁgure 1A). Strikingly, C57BL/6 mice pretreated with ﬂagellin before irradiation had signiﬁcantly reduced numbers of pycnotic or karyorrhexis cells compared with PBS-treated irradiated mice (ﬁgure 1A,B), indicating that ﬂagellin administration reduced the number of apoptotic cells in wildtype ilea. However, irradiated TLR5/ C57BL/6 mice treated with ﬂagellin showed no decrease in the number of pycnotic or karyorrhexis cells compared with the irradiated mice treated with PBS. In fact, irradiated TLR5/ C57BL/6 mice exhibited elevated crypt apoptosis when pretreated with ﬂagellin, compared with mice pretreated with PBS, indicating that ﬂagellin potentially has a radiosensitising effect on mice devoid of TLR5-mediated signalling (ﬁgure 1A,B). We also examined radiation-induced apoptosis in ilea of C57BL/6 or TLR5/ C57BL/6 mice using TUNEL assay, which directly detects nuclear DNA fragmentation 649

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon

Figure 1 Flagellin pretreatment reduces ionising radiation-induced apoptosis in the intestinal mucosa by a Toll-like receptor 5 (TLR5)-mediated mechanism. (A) H&E stain of caeca from C57BL/6 and C57BL/6 (tlr5/) mice administered 50 mg flagellin (intraperitoneally) 2 h before insult by 8 Gy of g-radiation. Arrows indicate pycnotic cells. (B) Quantitative representation of pycnotic cells per crypt in genotypes and treatments described in (A). Student t test (n¼40) (*p<0.05) (C) TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay analysis of caecal sections from C57BL/6 and C57BL/6 (tlr5/) mice treated as described in (A). Arrows indicate TUNEL-positive cells. (D) Quantitative representation of TUNELpositive cells in thge genotypes and treatments described in (C). Student t test (n¼40) (*p<0.05). (E) Immunostain analysis of caecal sections from C57BL/6 and C57BL/6 (tlr5/) mice treated as described in (A) with an antibody against cleaved caspase-3. Arrows indicate cleaved caspase-3positive cells. (F) Quantitative representation of caspase-3-positive cells in the genotypes and treatments described in (E). Student t test (n¼40) (*p<0.05).

650

Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon in late stage apoptotic cells. Consistent with data described above, C57BL/6 wild-type mice pretreated with ﬂagellin before irradiation had markedly decreased numbers of TUNEL-positive cells in ileal crypts, compared with control mice pretreated with PBS. However, ﬂagellin pretreatment before irradiation did not reduce the number of TUNEL-positive cells in the ileal cryps of TLR5/ C57BL/6 mice (ﬁgure 1C,D). Finally, ileal crypts of irradiated mice were examined by immunohistochemistry using an antibody against cleaved caspase-3. Again, wild-type C57BL/6 mice pretreated with ﬂagellin 2 h before irradiation had decreased numbers of active caspase-3-positive cells compared with isogenic controls treated with PBS, whereas TLR5/ C57BL/6 mice showed no evidence of decreased numbers of apoptotic cells following ﬂagellin pretreatment (ﬁgure 1E,F). Together, these data indicate that pretreatment of mice with ﬂagellin reduces radiation-induced apoptosis in the murine intestinal crypt cells compartment by a TLR5-dependent mechanism.

Flagellin pretreatment or constitutive TLR5 activation reduces ionising radiation-induced apoptosis in cultured intestinal epithelial cells (IEC-6) In order to determine the extent to which ﬂagellin mediates cytoprotective effects on epithelial cells, we subjected cultured epithelial cells to radiation insult following ﬂagellin pretreatment. Radiosensitive cultured IEC-6 were treated with ﬂagellin or PBS for 2 h before insult with 12 Gy of radiationdthe minimum radiation dose required to induce apoptosis in cultured IEC-6. Apoptotic cells entering late-phase programmed cell death were detected by TUNEL assay, and counted per 203

microscopy ﬁeld. Consistent with our previous observations in mouse epithelial mucosa, cultured cells pretreated with ﬂagellin had visibly fewer TUNEL-positive cells following irradiation compared with cells pretreated with PBS (ﬁgure 2A,B). We also stained these cell populations with annexin V and propodium iodide before analysis by ﬂow cytometry. Cell populations pretreated with ﬂagellin exhibited a marked decrease in annexin V-positive cells compared with irradiated cells or unirradiated cells treated with ﬂagellin alone (ﬁgure 2C,D). These results also indicate that the cytoprotective effects of ﬂagellin can be modelled in cultured IEC-6, and that the observed cytoprotective effect is epithelial, and not due to a paracrine effects.

Flagellin induces upregulation of cytoprotective genes in murine small intestine Following our observations of ﬂagellin-mediated protection from radiation-induced injury in murine radiosensitive intestinal tissues and cultured IEC-6, we undertook an expression proﬁling study of genes upregulated in the small intestine in response to systemic ﬂagellin. C57BL/6 mice were treated intraperitoneally with ﬂagellin or PBS for 2 h before preparation of total RNA from murine ilea. The abundance of mRNA species in the intestinal tissue was measured by cDNA microarray according to microarray system and data management methodology reported previously.15 Gene expression levels from experimental ﬂagellin-treated animals were normalised against the levels for untreated mice. Flagellin treatment resulted in the upregulation of genes which have been shown to function as proinﬂammatory mediators, microbicidals, antiapoptotic agents,

Figure 2 Flagellin pretreatment or constitutive Toll-like receptor 5 (TLR5) activation reduces ionizing radiation-induced apoptosis in cultured intestinal epithelial cells. (A) TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labelling) assay analysis of IEC-6 cultured cells pretreated with flagellin for 2 h before insult by 12 Gy of g-radiation. (B) TUNEL-positive cells per 203 microscopy field in the assay described in (A). Student t test (n¼40) (*p<0.05). (C) Flow cytometry analysis for annexin V-positive cells in IEC-6 cultured cells pretreated with flagellin for 2 h before insult by 12 Gy of gradiation. (D) Quantitative representation of annexin V-positive cells described in (C). Green fluorescent protein (GFP)-positive cells were gated and the percentage of annexin V/GFP-positive cells was quantified. Student t test (n¼40) (*p<0.05). Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

651

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon antioxidants and stress-response genes (table 1). Particularly notable is the robust (16-fold) upregulation of dual-speciﬁcity phosphatase 16 (DUSP16) (ﬁgure 3A). Also known as MKP-7, this protein functions to dephosphorylate phospho-JNK, thus acting as a negative feedback loop to restore JNK pathway signalling to tonic levels. Because radiation-induced apoptosis is mediated via JNK pathway activation,10e12 we hypothesised that ﬂagellin-mediated induction of cellular MKP-7 levels (which leads to the dephosphorylation of cytoplasmic JNK) suppresses radiation-induced apoptosis. In order to recapitulate these data in a cultured cell model, we treated IEC-6 with ﬂagellin and then measured MKP-7 transcript levels at intervals up to 6 h using quantitative PCR. In cultured cells, MKP-7 transcript levels increased up to 3 h after ﬂagellin stimulation and then decreased to levels similar to those of unstimulated cells at 4 h after ﬂagellin stimulation (ﬁgure 3B). In a parallel experiment, we analysed JNK pathway activity in IEC-6 treated with ﬂagellin at identical time points. Levels of phosphorylated JNK increased to a maximum at 15 min after ﬂagellin stimulation, but were undetectable at 2 h (ﬁgure 3C). The reduction of phosphorylated JNK species directly coincides with elevated MKP-7 levels up to 2 h poststimulation. Furthermore, elevated levels of MKP-7 were detected in cultured IEC-6 by immunoblot analysis up to 3 h after ﬂagellin stimulation (ﬁgure 3D). To conﬁrm that repression of MKP-7 functions in JNK dephosphorylation, siRNA against MKP-7 was transfected into IEC-6 for 48 h. A marked dosedependent increase in phosphorylated JNK was detected in these cells, showing that inhibition of MKP-7 translation results in constitutive activation of JNK signalling (ﬁgure 3E). Cells transfected with siRNA against MKP-7 also exhibited elevated levels of apoptosis even in the absence of external stimulation; levels that were further increased following irradiation (ﬁgure 3F). Together, these data show that MKP-7 is an inhibitor of JNK pathway signalling, and MKP-7 loss of function induces cellular apoptosis.

Constitutive MKP7 expression potently inhibits JNK phosphorylation and reduces radiation-induced apoptosis in cultured cells To determine the extent to which the cytoprotective effects of ﬂagellin pretreatment can be mediated by constitutive MKP-7 expression, we cloned the MKP-7 gene sequence into the

mammalian expression vector pCMV-myc. Additionally, we created a control construct harbouring a catalytically inactive form of MKP-7 (mMKP-7), where a critical residue known to be required for MKP-7 catalytic activity was mutated to an alanine residue. To test whether the catalytic activity was preserved in MKP-7, and to test whether the catalytic activity had been abolished in mMKP-7, we co-transfected these constructs, along with plasmids harbouring the JNK pathway intermediates JNK1 and MKK4 to initiate pathway activation, into cultured HEK293 cells. As expected, pCMV-myc-MKP-7 potently abrogate induced levels of phosphorylated JNK1, whereas pCMV-myc-mMKP-7 could not (ﬁgure 4A). Bacterial ﬂagellin is a known ligand for several pattern recognition receptors, including IPAF and TLR5. In order to determine the extent to which recombinant MKP-7 could abrogate phosphorylation of JNK in response to TLR5 activation, we investigated the effects of transfecting cultured cells with a plasmid harbouring a constitutively active form of TLR5 transcriptionally fused to the surface receptor cluster of differentiation 4 (CD4). We created the pEF6-V5-CD4TLR5 construct, which includes a CD4 extracellular domain on the N-terminus, and a TLR5 intracellular (including the TIR domain) domain on the C-terminus. Similar to the data described in ﬁgure 4A, pCMVmyc-MKP-7 potently abrogated induced levels of phosphorylated JNK1 initiated by pEF6-V5-CD4TLR5, whereas pCMV-mycmMKP-7 could not (ﬁgure 4B). In order to determine the extent to which constitutive MKP-7 expression abrogates apoptosis resulting from irradiation insult, we co-transfected cultured IEC-6 with pcDNA-EGFP (enhanced green ﬂuorescent protein) and pCMV-myc-MKP7 or pCMV-myc-mMKP7, respectively. After 24 h cells were subjected to 12 Gy irradiation and incubated for another 24 h. Cells were then stained with annexin V and propodium iodide and analysed by ﬂow cytometry. Data analysis where GFP-positive populations were gated and the percentage of annexin V-positive cells determined showed a decrease in the number of irradiation-induced apoptosis-positive cells in the population expressing MKP-7, but not in populations expressing mMKP-7 or vector alone (Supplementary ﬁgure 1). However, a caveat of using cultured IEC-6 in transfection studies is the relatively low transfection efﬁciency of this cell line; in the order of 5e10% efﬁciency. To achieve high efﬁciency expression of MKP7 in radiosensitive cells, we subcloned myc-MKP7 and mycmMKP7 into pShuttle-IERS-hrGFP-1 vector. These two

Table 1 Flagellin induces expression of cytoprotective molecules in the murine intestine: microarray analysis of mRNA isolated from murine intestine 2 h after intraperitoneal administration of 50 mg of flagellin

652

Fold induction

Gene

Common name

Function

14.7 12.0 6.8 6.0 5.4 4.1 4.0 3.4 3.1 2.9 2.8 2.8 2.6 2.6 2.2 2.1 2.0

Defcr3 MKP-7 Mmp17 Capn2 Itgb7 Irak4 Defcr23 Gpx7 Klf4 Mmp9 Gstm3 AK079781 Ciapin1 Klf5 Ocln Faim Defb37

Defensin-related cryptdin 3 MKP-7/DUSP16 Matrix metalloproteinase 17 Calpain 2 Integrin b 7 Interleukin-1 receptor-associated kinase 4 Defensin-related cryptdin 23 Glutathione peroxidase 7 Kruppel-like factor 4 Matrix metalloproteinase 9 Glutathione S-transferase Dual oxidase 2 Cytokine induced apoptosis inhibitor 1 Kruppel-like factor 5 Occludin Fas apoptotic inhibitory molecule 1 Defensin b 37

Antimicrobial JNK inhibition Tissue remodelling Cell proliferation differentiation Leucocyte homing Adaptor molecule Antimicrobial Antioxidant Cell proliferation Tissue remodelling Cell detoxification, chemoprotection ROS production Anti-apoptotic protein Cell proliferation Tight junction formation Antiapoptotic protein Antimicrobial

Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon Figure 3 Flagellin induces the expression of c-Jun N-terminal kinase (JNK)-specific phosphatase mitogenactivated protein kinase phosphatase 7 (MKP-7; DUSP16) in cultured IEC-6. (A) Quantitative PCR analysis of MKP-7 gene expression in murine ileum preparations at 2 h following intraperitoneal administration of 50 mg of flagellin. (B) Quantitative PCR analysis of MKP-7 gene expression in cultured IEC-6 after treatment with flagellin. (C) Immunoblot analysis of cultured IEC-6 treated with flagellin using an antibody against phosphorylated JNK. (D) Immunoblot analysis of lysates from cultured IEC-6 treated with flagellin for the indicated periods, using an antibody against MKP-7. MKP-7 was detected at 73 kDa. A non-specific (N.S.) band at 95 kDa was used as loading control. (E) Immunoblot analysis of lysates of IEC-6 either untreated (U) or transfected with non-targeting (NT) or MKP-7 small interfering RNA (siRNA), using an antibody against phosphorylated JNK. (F) Flow cytometry analysis for annexin V-positive cultured IEC-6 transfected with 10 nM non-targeting (NT) or MKP-7 siRNA. Some transfected cell populations were analysed before irradiation (Un) and others at 24 h following insult by 12 Gy of g-radiation (Irr.). Error bars indicate the SEM. GADPH, glyceraldehyde phosphate dehydrogenase. constructs or the pShuttle-IERS-hrGFP-1 vector alone was then recombined with pAdEasy-1 adenoviral vector to make pAdEasymyc-MKP7-GFP, pAdEasy-myc-mMKP7-GFP and pAdEasy-GFP. The three viral constructs were packaged, ampliﬁed and puriﬁed to a titre of 107 pfu/ml. Packaged adenovirus were then infected into cultured IEC-6 with an estimated efﬁciency of 80% as determined by counting GFP-positive cells (Supplementary ﬁgure 2). Importantly, and consistent with data presented in ﬁgure 3E using siRNA against MKP-7, infection with mutant and catalytically inactive pAdEasy-myc-mMKP7-GFP resulted in elevated levels of phosphorylated JNK at a multiplicity of infection (MOI) of >25, whereas no elevation in phosphorylation of JNK was detected at an MOI of 10 (ﬁgure 4C). We then conﬁrmed that the activity of MKP-7 expressed from the adenovirus at an MOI of 10 was sufﬁcient to inhibit ﬂagellin-induced JNK phosphorylation (ﬁgure 4D). A population of IEC-6 infected at an MOI of 10 with the aforementioned adenoviruses were subjected to 12 Gy irradiation, incubated for 24 h and stained with annexin V and propodium iodide before analysis by ﬂow cytometry. Consistent with the data described above, adenovirus-mediated expression of MKP-7 reduced the number of annexin V-positive cells within the GFP-positive IEC-6 population, showing that blockade of JNK pathway signalling during irradiation insult inhibits apoptosis and diverts cellular fate (ﬁgure 4E,F). Together, these data show that irradiation insult-induced apoptosis can be inhibited by constitutive MKP-7 expression.

Flagellin pretreatment inhibits ionising radiation-induced JNK phosphorylation and activation of apoptotic protein markers We then investigated the extent to which ﬂagellin pretreatment could inhibit radiation-induced JNK phosphorylation in cultured Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

IEC-6. Cells were treated with ﬂagellin or PBS for 2 h, then washed with Hanks medium before insult with 12 Gy of radiation. Thereafter, lysates of stimulated or control cells were collected at regular time intervals up to 6 h after ﬂagellin treatment and analysed by immunoblot. As expected, ﬂagellin induced JNK phosphorylation up to 1 h poststimulation, whereupon levels of phosphorylated JNK rapidly became undetectable at 2 h after ﬂagellin stimulation, presumably due to the phosphatase activity of MKP-7 (ﬁgure 5A). At 3 and 4 h after ﬂagellin treatment (1 and 2 h postradiation insult), we detected the presence of phosphorylated JNK1 (46 kDa) and, to a much lesser extent, phosphorylated JNK2 and JNK3 (both 54 kDa), only in PBS-treated cells but not in the ﬂagellin-treated cells (ﬁgure 5A). Thereafter, the elevation of the apoptotic marker cleaved PARP (poly(ADP-ribose) polymerase) was detected in PBS-treated irradiated cells up to 6 h (4 h postradiation insult) (ﬁgure 5A). Importantly, detectable levels of cleaved PARP were reduced in cells pretreated with ﬂagellin before radiation insult, although the emergence of a faint cleaved PARP band at 6 h in the ﬂagellin-treated irradiated cells was probably due to incomplete protection of the whole cell population within the cultured cell model system. To conﬁrm that MKP-7 activity reduces radiation-induced apoptosis, cultured IEC-6 were infected with adenovirus particles harbouring pAdEasy-myc-MKP-7-GFP, pAdEasy-myc-mMKP-7-GFP or pAdEasy-GFP. Cells were then incubated for 24 h, before being subjected to 12 Gy radiation insult. Thereafter, lysates were prepared from the treated cells and analysed by immunoblot. Consistent with the above data, phosphorylated JNK1 and, to a lesser extent, phosphorylated JNK2 and JNK3 were detected in pAdEasymyc-mMKP-7-GFP or pAdEasy-GFP cell lysates at 1 and 2 h 653

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon

Figure 4 Constitutive mitogen-activated protein kinase phosphatase 7 (MKP-7) expression potently inhibits c-Jun N-terminal kinase (JNK) phosphorylation and reduces radiation-induced apoptosis in cultured cells. (A) Immunoblot analysis of lystaes from 293HEK cultured cells cotransfected with plasmids harbouring myc-JNK1 and flag-MKK7 (in combination they constitutively activate the JNK pathway) and with plasmids harbouring myc-MKP7 and the catalytically inactive myc-mMKP7. Blots were probed using antibodies against phosphorylated JNK, against c-Myc for detection of JNK1, MKP-7 and mMKP-7 expression, against Flag for detection of MKK7 expression, and against b-actin as loading control. (B) Immunoblot analysis of lysates from 293HEK cultured cells co-transfected with plasmids harbouring V5-CD4-TLR5 and with plasmids harbouring mycMKP7 and myc-mMKP7. Blots were probed with the same antibodies used in (A) except for anti-V5 which was used to detect expression of CD4-TLR5. (C) Immunoblot analysis of cultured IEC-6 infected with adenovirus harbouring pAdEasy-GFP at a multiplicity of infection (MOI) of 100, of pAdEasymyc-tMKP-7-GFP at an MOI of 100 or pAdEasy-myc-mMKP-7-GFP at an MOI of 10, 25, 50 or 100, using an antibody against phosphorylated JNK. (D) Immunoblot analysis of cultured IEC-6 transduced with adenovirus harbouring pAdEasy-GFP, pAdEasy-myc-tMKP-7-GFP or pAdEasy-myc-mMKP-7GFP at an MOI of 10 and treated with flagellin for 15 min. (E) Flow cytometry analysis of cultured IEC-6 infected by adenovirus harbouring pAdEasyGFP, pAdEasy-myc-tMKP-7-GFP or pAdEasy-myc-mMKP-7-GFP at an MOI of 10, before insult by 12 Gy of g-radiation. (F) Quantitative representation of annexin V-positive cells detected in (E). Student t test (n¼9) (*p<0.05). GFP, green fluorescent protein.

654

Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon Figure 5 Flagellin pretreatment inhibits ionising radiation-induced c-Jun N-terminal kinase (JNK) phosphorylation and activation of apoptotic protein markers (A) Immunoblot analysis of cultured IEC-6 lysates pretreated for 2 h with either phosphate-buffered saline (PBS) (left-hand panels) or flagellin (right hand panels) and then subjected to insult by 12 Gy of g-radiation. Blots were probed using antibodies against phosphorylated JNK, and cleaved PARP (poly(ADP-ribose) polymerase). (B) Immunoblot analysis of cultured IEC-6 infected by adenovirus harbouring pAdEasy-GFP, pAdEasy-myc-tMKP-7GFP or pAdEasy-myc-mMKP-7-GFP before insult by 12 Gy of g-radiation and analysed by immunoblot using the same antibodies as described in (A). GFP, green fluorescent protein. postradiation insult, along with cleaved PARP at intervals up to 6 h (ﬁgure 5B). No phosphorylated JNK1 or cleaved PARP was detected in lysates from cells harbouring pAdEasy-myc-MKP-7GFP. Together, these data show that expression of MKP-7 inhibits radiation-induced JNK phosphorylation and PARP cleavage.

cells and enteroendocrine cells which, together with intercellular junctional proteins, form a tight biological barrier. Spontaneous or tonic levels of apoptosis occur in small intestine crypts and the luminal surface of healthy mammalian intestine as part of normal gut maintenance or as part of the homeostatic mechanism regulating stem cell numbers. However, elevation of

Flagellin pretreatment inhibits the intrinsic/mitochondrial apoptotic pathway in response to ionising radiation insult Apoptosis resulting from radiation insult activates the intrinsic/ mitochondrial apoptotic pathway.10 11 A number of proteins have been shown to be required for intrinsic/mitochondrial apoptotic pathway signalling, including Bcl-2 which is phosphorylated during intrinsic pathway activation, Bax that becomes a part of the mitochondrial membrane upon pathway activation, and cytochrome c, which is released from the mitochondria during apoptosis. To show that ﬂagellin pretreatment before radiation insult inhibits intrinsic pathway-induced apoptosis, we repeated experimental procedures described in ﬁgure 5, except that lysates were prepared from IEC-6 up to 2 days following irradiation. Lysates were analysed by immunoblot using antibodies against Bcl-2. In control cells subjected to radiation insult only, levels of Bcl-2 gradually decreased up to 2 days after irradiation (ﬁgure 6A). However, in ﬂagellin-pretreated irradiated cells, no alteration in Bcl-2 levels was detected (ﬁgure 6A). Then, we investigated cellular Bax and cytochrome c distribution in ﬂagellin-treated and irradiated IEC-6. Experiential lysates were collected, fractionated by centrifugation and analysed by immunoblot using anti-Bax or anti-cytochrome c antibodies. Consistent with the data presented above, Bax was recruited to the mitochondria up to 48 h postirradiation (ﬁgure 6B), whereas increased cytochrome c was detected in cystolic fractions of IEC-6 over the same period in cells subjected to radiation insult alone (ﬁgure 6C). No recruitment of Bax to the mitochondria or release of cytochrome c into the cytosol was detected when cells were pretreated with ﬂagellin before radiation insult. Thus, these data demonstrate that ﬂagellin pretreatment prior to irradiation inhibits the intrinsic cell death pathway.

DISCUSSION The intestinal epithelium is the most dynamically renewing tissue in adult mammals. Stem cells exiting the intestinal crypts towards the villus apex differentiate into enterocytes, goblet Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

Figure 6 Flagellin pretreatment inhibits the intrinsic/mitochondrial apoptotic pathway in response to ionising radiation insult. (A) Immunoblot analysis of cultured IEC-6 lysates pretreated with flagellin for 2 h before insult by 12 Gy of g-radiation. Blots were probed using antibodies against Bcl-2. (B) Immunoblot analysis of the mitochondrial fraction of lysates of cultured IEC-6 pretreated with flagellin for 2 h and then subjected to insult by 12 Gy of g-radiation using an antibody against Bax (C) Immunoblot analysis of the cyotplasmic fraction of lysates of cultured IEC-6 pretreated with flagellin for 2 h and then subjected to insult by 12 Gy of g-radiation using an antibody against cytochrome c (CytC). Hsp60, heat shock protein 60; PBS, phosphate-buffered saline. 655

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon normal gut apoptosis leads to gastrointestinal disease, a condition characteristic of many enteric infections and commonly manifested in patients with cancer during irradiation treatment. Ionising radiation considerably elevates normal levels of apoptosis, particularly in the small intestine crypts at 2e6 h following exposure. Elevated apoptosis directly leads to reduced enterocyte numbers and weakened barrier function, which potentially allow resident ﬂora to invade intestinal organs and tissue. Strategies to limit ionising radiation-induced gut apoptosis have not been substantially explored. Here, we describe the specialised use of ﬂagellin as potential adjuvant therapy for limiting irradiation-induced gut apoptosis and identify MKP-7 as a potent cytoprotective factor transiently upregulated in response to ﬂagellin. Our investigations revealed that MKP-7 expression is markedly elevated in the murine small intestine in response to ﬂagellin at 2 h post-treatment. MKP-7 functions in the selective inhibition of the JNK pathway by potently dephosphorylating p-JNK, with highly conserved functionality.17 19 22 Consistently, constitutive MKP-7 expression in cultured cells potently inhibited the phosphorylation of JNK, thereby also inhibiting cellular outcomes of JNK pathway activation, including ionising radiation-induced apoptosis. Due to an MKP-7-mediated negative feedback loop, p-JNK is undetectable at 2 h after ﬂagellin treatment, a time point that directly correlates with maximal MKP-7 expression. Hence, the optimal time for ﬂagellin administration is at a time window 2 h before ionising radiation insult. Flagellin pretreatment outside of this time window failed to reduce ionising radiation-induced apoptosis in the murine small intestine. As well as the radioprotective effects from the modulation of JNK signalling reported in this study, other reports have also demonstrated the radioprotective effects of the nuclear factor-kB (NF-kB) pathway.4 23 24 Interestingly, our expression proﬁling studies of mouse intestinal tissue only detected the twofold induction of the NF-kB-associated Faim antiapoptotic factor, and the most highly expressed NF-kB-induced genes detected were of reported antimicrobial function. Nevertheless, due to increased reports of cross-talk in the regulation of the NF-kB and JNK pathways,25 it is highly plausible that proteins upregulated by either pathways elicit temporal cytoprotective effects. Here, we report on the potential for utilising puriﬁed ﬂagellin as an adjuvant cytoprotective therapy within a distinct time frame (2 h) before irradiation. Recent reports have shown that recognition of commensal microﬂora by TLRs is required for intestinal homeostasis.26 However, disease prevention approaches where innate immunity is activated by bacterial products or proteins have been given little consideration due to the potential for severe inﬂammatory pathology. Nevertheless, not all bacterially derived innate immune activators potentiate equivalent inﬂammatory effects. For example, ﬂagellin has been shown to elicit comparatively mild pathological outcomes compared with lipolysaccharide administration, which results in severe inﬂammation and shock. The disparate pathological effect of bacterial products within the host is due to diverse cellular responsiveness to ﬂagellin. Whereas epithelial cells robustly respond to ﬂagellin,27e29 macrophages are devoid of TLR5,30 and thus ﬂagellin is a poor inducer of tumour necrosis factor a in mice.9 30 In fact, most populations of murine macrophages or dendritic cells do not respond to puriﬁed soluble ﬂagellin,31 32 and the majority of the ﬂagellin-induced cytokine elevation is mediated by MyD88dependent signalling in non-haemopoietic cells.33 Another noteworthy observation supporting the suitability of ﬂagellin pretreatment as adjuvant therapy is that ﬂagellin has a relatively 656

short half-life in mice. Flagellin is undetectable in the murine serum 2 h following intraperitoneal administration of 50 mg, suggesting efﬁcient immune mechanisms of proteolytic elimination, and also avoiding the possible determinative effect of prolonged signalling. Together, these data strongly support the use of puriﬁed soluble ﬂagellin as adjuvant therapy to reduce the harmful side effects of ionising irradiation-induced gut apoptosis. Strategies to activate antiapoptotic factors transiently with ﬂagellin pretreatment, a comparatively less toxic bacterial product, should be considered for development in specialised use in limiting determinative effects of external insult. Funding This work was supported in part by National Institutes of Health grants DK-71604 and AI-64462 (ASN), and DK-64399. Competing interests None. Contributors RMJ and ASN were involved in study concept and design and drafting of the manuscript. VS, HW, LL and AK were involved in acquisition of data. MVK and AG provided material support. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.

Potten CS, Grant HK. The relationship between ionizing radiation-induced apoptosis and stem cells in the small and large intestine. Br J Cancer 1998;78:993e1003. Hickman JA, Potten CS, Merritt AJ, et al. Apoptosis and cancer chemotherapy. Philos Trans R Soc B: Biol Sci 1994;345:319e25. Potten CS, Merritt A, Hickman J, et al. Characterization of radiation-induced apoptosis in the small intestine and its biological implications. Int J Radiat Biol 1994;65:71e8. Egan LJ, Eckmann L, Greten FR, et al. IkappaB-kinasebeta-dependent NF-kappaB activation provides radioprotection to the intestinal epithelium. Proc Natl Acad Sci USA 2004;101:2452e7. Neta R. Modulation of radiation damage by cytokines. Stem Cells 1997;15(Suppl 2):87e94. Riehl T, Cohn S, Tessner T, et al. Lipopolysaccharide is radioprotective in the mouse intestine through a prostaglandin-mediated mechanism. Gastroenterology 2000;118:1106e16. Abreu MT. Harnessing the power of bacteria to protect the gut. N Engl J Med 2008;359:756e9. Burdelya LG, Krivokrysenko VI, Tallant TC, et al. An agonist of toll-like receptor 5 has radioprotective activity in mouse and primate models. Science 2008;320:226e30. Vijay-Kumar M, Aitken JD, Sanders CJ, et al. Flagellin treatment protects against chemicals, bacteria, viruses, and radiation. J Immunol 2008;180:8280e5. Kanzawa T, Iwado E, Aoki H, et al. Ionizing radiation induces apoptosis and inhibits neuronal differentiation in rat neural stem cells via the c-Jun NH2-terminal kinase (JNK) pathway. Oncogene 2006;25:3638e48. Kim MJ, Lee KH, Lee SJ. Ionizing radiation utilizes c-Jun N-terminal kinase for amplification of mitochondrial apoptotic cell death in human cervical cancer cells. FEBS J 2008;275:2096e108. Dent P, Yacoub A, Fisher PB, et al. MAPK pathways in radiation responses. Oncogene 2003;22:5885e96. Kim MJ, Choi SY, Park IC, et al. Opposing roles of c-Jun NH2-terminal kinase and p38 mitogen-activated protein kinase in the cellular response to ionizing radiation in human cervical cancer cells. Mol Cancer Res 2008;6:1718e31. Weston CR, Davis RJ. The JNK signal transduction pathway. Curr Opin Cell Biol 2007;19:142e9. Zeng H, Carlson AQ, Guo Y, et al. Flagellin is the major proinflammatory determinant of enteropathogenic Salmonella. J Immunol 2003;171:3668e74. Keyse SM. Protein phosphatases and the regulation of mitogen-activated protein kinase signalling. Curr Opin Cell Biol 2000;12:186e92. Tanoue T, Yamamoto T, Maeda R, et al. A novel MAPK phosphatase MKP-7 acts preferentially on JNK/SAPK and p38 alpha and beta MAPKs. J Biol Chem 2001;276:26629e39. Katagiri C, Masuda K, Urano T, et al. Phosphorylation of Ser-446 determines stability of MKP-7. J Biol Chem 2005;280:14716e22. Masuda K, Shima H, Watanabe M, et al. MKP-7, a novel mitogen-activated protein kinase phosphatase, functions as a shuttle protein. J Biol Chem 2001;276:39002e11. Kamata H, Honda S, Maeda S, et al. Reactive oxygen species promote TNFalpha-induced death and sustained JNK activation by inhibiting MAP kinase phosphatases. Cell 2005;120:649e61. Vijay-Kumar M, Sanders CJ, Taylor RT, et al. Deletion of TLR5 results in spontaneous colitis in mice. J Clin Invest 2007;117:3909e21. McEwen DG, Peifer M. Puckered, a Drosophila MAPK phosphatase, ensures cell viability by antagonizing JNK-induced apoptosis. Development 2005;132:3935e46. Rasoulpour RJ, Boekelheide K. NF-kappaB activation elicited by ionizing radiation is proapoptotic in testis. Biol Reprod 2007;76:279e85.

Gut 2011;60:648e657. doi:10.1136/gut.2010.223891

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon 24. 25. 26. 27. 28.

Spehlmann ME, Eckmann L. Nuclear factor-kappa B in intestinal protection and destruction. Curr Opin Gastroenterol 2009;25:92e9. Liu J, Lin A. Wiring the cell signaling circuitry by the NF-kappa B and JNK1 crosstalk and its applications in human diseases. Oncogene 2007;26:3267e78. Rakoff-Nahoum S, Paglino J, Eslami-Varzaneh F, et al. Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis. Cell 2004;118:229e41. Gewirtz AT, Simon PO Jr, Schmitt CK, et al. Salmonella typhimurium translocates flagellin across intestinal epithelia, inducing a proinflammatory response. J Clin Invest 2001;107:99e109. Hawn TR, Verbon A, Lettinga KD, et al. A common dominant TLR5 stop codon polymorphism abolishes flagellin signaling and is associated with susceptibility to legionnaires’ disease. J Exp Med 2003;198:1563e72.

29. 30. 31. 32. 33.

Reed KA, Hobert ME, Kolenda CE, et al. The Salmonella typhimurium flagellar basal body protein FliE is required for flagellin production and to induce a proinflammatory response in epithelial cells. J Biol Chem 2002;277:13346e53. Miao EA, Alpuche-Aranda CM, Dors M, et al. Cytoplasmic flagellin activates caspase-1 and secretion of interleukin 1beta via Ipaf. Nat Immunol 2006;7:569e75. Means TK, Hayashi F, Smith KD, et al. The Toll-like receptor 5 stimulus bacterial flagellin induces maturation and chemokine production in human dendritic cells. J Immunol 2003;170:5165e75. Uematsu S, Jang MH, Chevrier N, et al. Detection of pathogenic intestinal bacteria by Toll-like receptor 5 on intestinal CD11c+ lamina propria cells. Nat Immunol 2006;7:868e74. Sanders CJ, Moore DA 3rd, Williams IR, et al. Both radioresistant and hemopoietic cells promote innate and adaptive immune responses to flagellin. J Immunol 2008;180:7184e92.

Editor’s quiz: GI snapshot A case of dyspepsia and abdominal fullness CLINICAL PRESENTATION A 59-year-old woman presented with a 5-month history of acid reﬂux symptoms which were initially relieved with a proton pump inhibitor. She complained of no weight loss, anorexia or abdominal pain but could feel a ‘fullness’ in her upper abdomen which prevented her from lying prone. Examination revealed a hard epigastric mass which seemed separate from the liver and spleen. There was no lymphadenopathy or other positive ﬁndings. Initial laboratory investigations showed a normal full blood count, urea and electrolytes, bilirubin and liver enzymes. Tumour markers revealed a carbohydrate antigen (CA-125) of 215 kU/l (reference range 0e35), CA-153 of 75 kU/l (0e35) and carcinoembryonic antigen (CEA) <1 mg/l (0e3). Upper gastrointestinal endoscopy showed extrinsic compression of the stomach but was otherwise unremarkable. A CT scan of the abdomen and pelvis was performed with intravenous and oral contrast (see ﬁgures 1 and 2). This demonstrated a ﬂuid-density mass within the upper abdomen.

Figure 2 Parasagittal reformat of the CT taken left of the midline demonstrating the maximal cranio-caudal dimensions of the fluid-density structure (asterisk).

QUESTION What is the nature and aetiology of the mass shown and what clinical syndrome has resulted? See page 694 for the answer

James D Thomas,1 Tanya M Monaghan,2 Jonathan James3 1

Department of Radiology, Nottingham University Hospitals NHS Trust, Nottingham, UK; 2Nottingham Digestive Diseases Centre, Biomedical Research Unit, Nottingham University Hospitals NHS Trust, Nottingham, UK; 3Nottingham Breast Institute, Nottingham University Hospitals NHS Trust, Nottingham, UK Correspondence to Dr James D Thomas, Department of Radiology, B Floor, Queen’s Medical Centre, Nottingham NG7 2UH, UK; [email protected] Competing interests None. Patient consent Obtained. Provenance and peer review Not commissioned; externally peer reviewed.

Figure 1 Axial image from the CT examination just inferior to the lower costal margin at the level of the renal veins. Gut May 2011 Vol 60 No 5

Published Online First 30 August 2010 Gut 2011;60:657. doi:10.1136/gut.2009.190264 657

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon

Accuracy of computed tomographic colonography in a nationwide multicentre trial, and its relation to radiologist expertise D Heresbach,1,2 M Djabbari,3 F Riou,4 C Marcus,5 A Le Sidaner,6 M A Pierredon-Foulogne,7 T Ponchon,8 M Boudiaf,9 J A Seyrig,10 H Laumonier,11 D Luet,12 M Giraud-Cohen,13 A L Pelletier,14 A Charachon,15 F Ramaholimihaso,16 P Bouillet,17 M Veyrac,18 S Ficarelli,19 K Vahedi,20 J Keruhel,21 H Lamouliatte,22 C Ridereau-Zins,23 Y Bouhnik,24 M Tissier,25 B Diris,26 A M Zagdanski,27 J M Josselin,2 S Hamonic,4 Y Gandon28 < Additional appendices are

published online only. To view these files please visit the journal online (http://gut.bmj. com). For numbered affiliations see end of article. Correspondence to Professor D Heresbach, Centre Hospitalier de Cannes, Unite´ D’Endoscopie Digestive, 15 avenue des brousaillles, 06414 Cannes cedex, France; [email protected] For other participants see end of the article Accepted 22 November 2010 Published Online First 25 January 2011

ABSTRACT Objective Reports on the accuracy of computed tomographic colonography (CTC) mainly involve series from expert institutions. The aims of this study were to assess CTC accuracy in a nationwide population and to relate it to radiologist performance in their initial training. Design Nationwide multicentre trial. Setting Twenty-eight radiologists, working in 26 mostly academic clinical units, were involved in the study after having attended a formal specialised 2-day training session on CTC. They worked through a training set of 52 cases with automatic feedback after an attempt at each case. Patients The study enrolled 845 patients with average and high risk of colorectal cancer, 737 of whom had both complete CTC and videocolonoscopy data, which constituted the dataset. Interventions Patients underwent same-day CTC followed by videocolonoscopy with segmental unblinding of CTC results. Main outcome measures Sensitivity, specificity and positive and negative predictive values for detection of polyps $6 mm in per-patient and per-lesion analyses of CTC without computer-aided detection. Results Sensitivity, specificity and positive and negative predictive values for patients with polyps $6 mm were 69% (95% CI 61% to 77%), 91% (95% CI 89% to 94%), 67% (95% CI 59% to 74%) and 92% (95% CI 90% to 94%), respectively. Univariate analysis showed that the detection rate for polyps $6 mm was linked to neither radiologist case volume nor number of polyps, but was related to sensitivity achieved in the training set. Pooled sensitivity was 72% (95% CI 63% to 80%) versus 51% (95% CI 40% to 60%) for radiologists achieving above and below median sensitivity in the training set (61%), respectively. Multivariate analysis showed that sensitivity for polyps $6 mm in the training set was the only remaining significant predictive factor for subsequent performance. Conclusions Radiologist sensitivity CTC for detection of polyps $6 mm in training was the sole independent predictor for subsequent sensitivity in detection of such polyps.

INTRODUCTION Computed tomographic colonography (CTC) is now available as a minimally invasive procedure for 658

detection of colonic polyps or early neoplasia. The efﬁcacy of CTC in detecting asymptomatic colorectal lesions is well established, but there is still debate about its use: in the USA the Medicaid and Medicare programmes do not cover CTC costs. The clinical evidence involves different size thresholds for polyp detection, heterogeneous study populations (with various risk levels for colorectal cancer (CRC)), and variable expertise among radiologists.1e3 The disparities in the accuracy of CTC limit its widespread use, although some training and credentialling programmes have been described. Few data have been published on the part played by radiologist expertise in predicting CTC accuracy in multicentre nationwide studies.4e6 Our multicentre nationwide study was designed to assess the accuracy of CTC in detecting small and large polyps or cancers ($6 mm in diameter) after a preliminary training and qualifying programme for radiologists. The lesions detected at CTC were conﬁrmed by videocolonoscopy, and the ﬁndings from the segmental unblinded videocolonoscopy after CTC data discovery and comparison were used as the reference standard as previously described.9

METHODS Patients A total of 26 clinical units involving 28 radiologists participated in the study, with approval obtained from their institutional review boards and from the French national agency responsible for the safety of health and medical products (AFSSAPS (Agence Française de Sécurité Sanitaire des Produits de Santé)). Participants were consecutively recruited between January 2007 and April 2008, from all those who attended participating centres to undergo routine videocolonoscopy for ﬁrst-line polyps or cancer screening, because they had symptoms, for surveillance because of family history or positive faecal occult bleeding test (FOBT), or for follow-up surveillance. Individuals were excluded from the study if they had inﬂammatory bowel disease, known familial polyposis, history of segmental colectomy, recent follow-up optical colonoscopy in the preceding 3 years, or if any comorbidity made optical Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon colonoscopy unsuitable. Each study participant provided written informed consent before enrolment. The total number of participants enrolled was 845. Complete CTC and complete videocolonoscopy results were available for 737 (87%). However, in three participants, the CRC risk was described as ‘indeﬁnite’ (ﬁgure 1), and these were excluded from the analysis in which CRC risk was the dependent variable; thus, in this case, 734 participants constituted the study dataset. A total of 303 participants (41%) had no known risk factors for CRC other than age. Of these, 50 (50/737; 7%) had a recent positive FOBT. All others (n¼253) were considered to be at average risk of CRC, although 202 (27%) had abdominal symptoms. The 431 participants at increased risk of CRC included 258 (258/737; 35%) who had a ﬁrst-degree relative with a history of colorectal adenoma or cancer and 173 (23%) who had a personal history of polyps or cancer. Demographic data are provided in table 1; the baseline demographic characteristics of the ﬁnal cohort were similar in each of the three subgroups.

of short lectures and hands-on workstation teaching. Each workstation equipped with CTC viewing software supplemented by CAD (Colon CAR 1.3; Medicsight PLC, Kensington Centre, England and Wales) was shared by two observers. Training was conducted by three radiologists with experience of over 300 validated CTC cases. Two of them are faculty members of the European Society of Gastrointestinal and Abdominal Radiology (ESGAR) hands-on CTC training workshops. Participants then had to work through a training set of 52 cases, with feedback being given to the trainee after their attempt at each case. This was done using a website speciﬁcally built for this training and study. In each case, after the participant has validated the results for the current case, it provides the number of lesions, with their histological type and descriptive ﬁndings resulting from a colonoscopy. For each lesion, participants had access to a set of CTC images and videocolonoscopic correlation data to allow them to match the lesions they had detected with the veriﬁed lesions. Details, characteristics and results of this training process and the 52-case training set are available online as appendix 1.

Radiologist training Each participating radiologist was required to conﬁrm previous expertise in abdominal CT and to attend one of the two formal specialised 2-day training sessions on CTC. Each session consisted

CTC and videocolonoscopy All participants underwent CTC followed by videocolonoscopy examination on the same day.

Cohort studied 845 participants from 26 centers 108 exclusions: Colonoscopy not done: 34 Incomplete colonoscopy: 37

737 included (384 men, 353 women)

3 indefinite CRC risk

367 without polyps

253 average risk (with symptoms or firstline endoscopic screening)

223 with polyps ≤ 5 mm

50 with recent positive FOBT

431 elevated risk Familial history of CRC: 215 Personal history of CRC or adenoma: 173

147 with at least 1 polyp ≥ 6 mm At least 1 adenoma or adenocarcinoma:120 At least 1 hyperplastic polyp: 21 At least 1 polyp of other histological type:6 214 polyps Hyperplastic polyps or polyp of other histological type:42 Adenomas or adenocarcinomas:172

Figure 1 Enrolment of participants in the study, colorectal cancer (CRC) risk subgroups, and polyp findings at computed tomographic colonoscopy (CTC). FOBT, faecal occult blood test. Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

659

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon Table 1 Characteristics of patients and numbers of patients with various histological types and lesion sizes, as detected by the reference methods, overall and according to colorectal cancer (CRC) risk All-risk groups (n[737)

Average CRC risk (n[253)

Elevated CRC risk (n[431)

Age at enrolment, years Mean (SD) 57.4 (0.4) 57.2 (0.7) 57.2 (0.5) IQR 51e65 50e67 51e65 Sex, n (%) Male 384 (52)y 131 (51.8) 219 (50.8) Female 353 (48)z 122 (48.2) 212 (49.2) Patients with at least one polyp of given size, n (% of risk group) Any size 370 (50.0)z 100 (39.5) 235 (54.5) #5 mm 317 (43.0)z 81 (32.0) 206 (47.8) >6 mm 147 (19.9)z 44 (17.4) 86 (20.0) $10 mm 70 (9.5) 20 (7.9) 37 (8.6) 6e9 mm 91 (12.3) 29 (11.5) 57 (13.2) $3 polyps <10 mm 99 (13.4) 18 (7.1) 72 (16.7) Patients with at least one adenoma of given size, n (% of risk group) Any size 244 (33.1) 66 (26.1) 153 (35.5) #5 mm 186 (25.2) 48 (19.0) 120 (27.8) $ 6 mm 120 (16.3) 34 (13.4) 69 (16) $10 mm 63 (8.5) 18 (7.1)* 32 (7.4)* 6e9 mm 68 (9.2) 21 (8.3) 42 (9.7) $3 adenomas <10 mm 47 (6.4) 12 (4.7) 30 (7.0) Patients with largest adenoma of given size, n (% of risk group) #5 mm 124 (16.8) 32 (12.6) 84 (19.5) 6e9 mm 56 (7.6) 16 (6.3) 37 (8.6) $10 mm 63 (8.5) 18 (7.1)* 32 (7.4)* Number of adenomas per patient, mean (SD) Any size 2.1 (0.11) 2.1 (0.22) 2.1 (0.15) $10 mm 0.36 (0.05) 0.42 (0.1) 0.30 (0.06) Patients with colorectal 7 (0.9) 4 (1.6) 1 (0.2) cancer, n (% of risk group)

Positive FOBT (n[50) 59.8 (1.1) 54e65 32 (64) 18 (36) 34 (68) 29 (58) 16 (32) 13 (26) 4 (8) 8 (16) 24 (48) 17 (34) 16 (32) 13 (26)* 4 (8) 4 (8) 8 (16) 3 (6) 13 (26)* 2.2 (0.26) 0.63 (0.13) 2 (4)

*p<0.05 for positive FOBT versus two other groups (c2 test and Fisher exact test). yTwo patients with indefinite risk. zOne patient with indefinite risk. FOBT, faecal occult blood test.

Patient preparation Preparation for CTC included stool tagging, laxative purgation and ﬂuid tagging, as previously described.7 8 Patients had a clear liquid diet during the preceding 24 h, with intake of 30 ml sodium phosphate and 20 mg bisacodyl the day before the CTC. Participants also ingested 500 ml barium sulfate (2.1% by weight) for solid stool tagging and 100 ml diatrizoate meglumine for opaciﬁcation of luminal ﬂuid. On the morning of the two consecutive procedures, 10 mg bisacodyl was administered anally. Colonic insufﬂation was attained with an automated carbon dioxide insufﬂator (PROTOCO2L; E-Z-EM, Bracco Imaging S.p.A. Milano, Italy). Manual insufﬂation with room air was used if adequate colon distension could not be obtained with the mechanical insufﬂator.

Computed tomographic colonography CTC data were obtained with patients in both the supine and prone positions as previously described9 and detailed in appendix 6 (online). To compare the results of CTC with videocolonoscopy, CTC analysis was performed without computer-aided detection.

Optical colonoscopy After the CTC examination, videocolonoscopy was performed according to the standard clinical protocol of each participating 660

site and with segmental unblinding of the CTC results.9 All optical colonoscopy examinations were performed by an experienced endoscopist who had carried out at least 500 complete colonoscopies. The colonoscopes were those used in daily practice; they usually had high deﬁnition, but high magniﬁcation was not available as other adjunctive techniques for polyp diagnosis. All polyps were categorised according to the Paris classiﬁcation.10 Tissue samples from all lesions were analysed by an experienced gastrointestinal pathologist and described according to the Vienna classiﬁcation.11 These data provided the reference standard for this study. In accordance with previous studies, the reference lesion size was determined endoscopically and was estimated using opened biopsy forceps. Quality of imaging was recorded by the radiologist, and overall results were described using the CT Colonography Reporting and Data System (C-RADS).12 The quality of bowel preparation was graded by the endoscopist using a ﬁve-category scale (‘excellent’ to ‘not interpretable’).13 Withdrawal times were not included, as these data were not routinely available from the optical colonoscopy reports at the time of the inclusions.

Definitions The results of segmental unblinded colonoscopy (including a second colonoscopy, when performed) and pathological ﬁndings for tissue specimens were the reference standard for lesion size, location and histological type. Like other recent prospective CTC screening studies,14 15 our study focused on lesions that were $6 mm, since the prevalence of advanced histological features in diminutive polyps (ie, #6 mm) is reportedly below 2% and carcinomas are rarely described in such small lesions. A positive CTC result was deﬁned as identiﬁcation of a lesion measuring $6 mm in diameter as previously described.16 If CTC was positive, and examinations identiﬁed one or more lesions over a given size threshold (eg, any lesion $6 mm or any lesion $10 mm) that was comparable and located in the same segment (with segments deﬁned as caecum, ascending, transverse, descending and sigmoid colon, or rectum) as the reference standard, the CTC result was considered to be a true positive with regard to lesions over that size threshold. If CTC was positive but no lesions over a given size threshold were found using the reference standard, the CTC result was considered to be false positive for lesions over that size threshold in that location. The per-patient accuracy of CTC was deﬁned as the proportion of patients well categorised by both the index test and the reference standard.

Data analysis The per-lesion sensitivity and positive predictive value (PPV), as well as per-patient sensitivity, speciﬁcity, PPV, and negative predictive value (NPV), were calculated with data pooled from all the radiologists. All other quantitative variables are expressed as means and standard deviations, or medians, and qualitative variables as numbers and percentages. The radiologisteendoscopist agreement for measures of polyp diameter was examined using the BlandeAltman method, with calculations of the mean difference 61 and 61.96 standard deviations. This agreement was also analysed using the k method, where polyps were then classiﬁed into three categories: #5 mm, 6e9 mm and $10 mm. Distributions of qualitative variables were compared, between independent groups using the c2 test or Fisher exact test, and between paired groups using the McNemar c2 test. The signiﬁcance threshold was set at p¼5%. Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon Radiologists’ sensitivities were analysed, through bivariate analysis, according to patients’ level of CRC risk and according to four radiologist variables: radiologist polyp detection rate with the initial training set for lesions <6 mm, and for lesions $6 mm, number of patients included in the study (in other words the radiologist’s case volume), and number of polyps that should have been detected according to the reference in each radiologist’s cases. For analysis involving each of these four variables, radiologists were divided into two subgroups according to the median value of the variable. Sensitivities to be compared were set, per-lesion and perpatient, for polyps $6 mm, polyps $10 mm, adenomas $6 mm and adenomas $10 mm. We then carried out a multiple logistic regression in order to identify variables linked to the dependent variabledthat is, radiologist detection rate for polyps $6 mm. The variables introduced as continuous ones (calculated for each radiologist) were sensitivity for polyp diagnosis shown with the training set, number of patients examined, number of polyps in the patients examined by each radiologist, and those variables described in other studies to be associated with diagnostic sensitivity for polypsdthat is, proportion of patients with a positive FOBT, proportion of patients with excellent, good or moderate large-bowel preparation, and proportion of ﬂat polyps. All analyses were carried out using the SAS V.9.1 software.

CTC performance Per-patient analysis The CTC performance for detecting patients with at least one polyp or one adenoma measuring >6 mm in diameter as well as lesions with a diameter of 6e9 mm, and patients with at least $3 lesions 6e9 mm in diameter is shown in table 3. Table 3 also shows the sensitivity, PPV, speciﬁcity and NPV for identifying patients with at least one polyp or one adenoma measuring $10 mm in diameter. Per-patient estimates of CTC accuracies for subgroups of patients with various CRC risk levels are shown in table 3. With regard to the nine carcinomas in seven patients, CTC detected eight out of nine cancers, identifying six out of seven patients. The undetected carcinoma presented in a patient whose examination was classiﬁed by the radiologist as C0 (ie, conﬁdent interpretation of the colonic ﬁndings was not possible because of, eg, incomplete bowel preparation or insufﬁcient bowel insufﬂation); in this case the cause was insufﬁcient bowel insufﬂation.

Per-lesion analysis

The sensitivity and PPV of CTC for lesions $6 mm or of size 10 mm or more as well as for lesions with 6e9 mm diameter are described in appendix 2 (online only). The CTC sensitivity for CRC was 67.6%. CTC detected, as mentioned, eight of the nine carcinomas.

RESULTS Characteristics of patients and lesions

CTC sensitivities according to radiologist characteristics Univariate analyses

Table 1 shows patient characteristics according to CRC risk level, and the distribution among them of various histological types and lesion sizes, found by the reference methods. Table 2 shows the distribution among patients of lesions according to size and shape using the Paris classiﬁcation. A total of 214 polyps measuring $6 mm in diameter were detected in 147 participants. A total of nine carcinomas (Vienna classiﬁcation 4.5 or higher) were detected in seven patients; the macroscopic shape was stenosis or tumour in ﬁve cases, ﬂat in three, and sessile in one, and the mean size was 37.2 mm (median 40, range 20e60). Patients with a recent positive FOBT were described separately because, although these patients were basically considered to be at moderate CRC risk before the FOBT was performed, a positive FOBT result is associated with an increased prevalence of polyps $10 mm. Effectively, in the present study, the prevalence of polyps with a diameter $10 mm was higher for patients with a recent positive FOBT than for the others.

As mentioned, we investigated the effect of radiologist characteristics by dividing the radiologists into two subgroups, according to the median of the distribution of the variable of interest, and comparing the sensitivities of the subgroups. The pooled CTC sensitivities for detecting polyps or adenomas $6 mm or $10 mm, according to initial performance (ie, detection rate for polyps of any size and for polyps $6 mm, in the initial training set) and experience (ie, number of patients enrolled in the study for each radiologist, and the number of polyps they should have detected in their own cases) are shown in table 4 (per-patient analysis). Similar results were found when the measure for initial performance was detection for polyps of any size in the initial training set. The per-lesion analysis data are available online in appendix 3. Sensitivity for detecting polyps or adenomas $6 mm at CTC was found to be signiﬁcantly higher in the per-lesion analysis for the subgroup of radiologists who had achieved the higher polyp detection rate during training, for the subgroup who had examined the higher numbers of patients, and for the subgroup with the higher number of polyps detected by the reference methods. The differences were not signiﬁcant when the analysis was performed on a per-patient basis except for diagnosis of adenomas $6 mm by the subgroups of radiologists with better performance in the training set or diagnosis of adenomas >10 mm by radiologists with a higher case volume (table 4).

Table 2 Patients and lesions, according to diameter and macroscopic shape of lesions as detected by the reference methods Polyps

Adenomas

Lesions n (%)

Patients n (%)

Total 895 Diameter #5 mm 681 (76) $ 6 mm 214 (24) $10 mm 99 (11) 6e9 mm 115 (13) $3 polyps with diameter <10 mm NA Macroscopic shape (Paris classification) 0-I p (pedunculate) 72 (8) 0-I s (sessile) 583 (65) 0-II (flat) 225 (25) Stenosis or tumour 5 (0.6)

370

518

(86) (40) (19) (25) (27)

346 (67) 172 (33) 88 (17) 84 (16) NA

186 120 63 68 47

57 (15) 265 (72) 132 (36) 5 (1.4)

62 (12) 346 (67) 104 (20) 5 (1.0)

50 (20) 181 (74) 61 (25) 5 (2.0)

317 147 70 91 99

Values are number (%).

Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

Lesions n (%)

Patients n (%) 244 (76) (49) (26) (28) (19)

Multivariate analysis Finally, we carried out a multivariate analysis by multiple logistic regression that included all the variables known to be signiﬁcantly associated in the univariate analysis with diagnostic sensitivity for polyps $6 mm: sensitivity for polyp diagnosis shown with the training set, number of patients examined, and the number of polyps in the patients examined by each radiologist, the proportion of patients with a positive FOBT, proportion of patients with excellent, good or moderate large-bowel preparation, and the proportion of ﬂat polyps in each subgroup of patients examined by each radiologist. 661

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon Table 3 Diagnostic value results for computed tomographic colonoscopy (CTC) in per-patient analysis with segmental unblinded colonoscopy and histology as reference standard, according to colorectal cancer (CRC) risk Per-patient at average risk (n¼253) Overall Polyps $6 mm Polyps $10 mm Polyps 6e9 mm $3 polyps with diameter <10 mm Adenomas $6 mm Adenomas $10 mm Adenomas 6e9 mm $3 adenomas with diameter <10 mm Per-patient at elevated risk (n¼431) Overall Polyps $6 mm Polyps $10 mm Polyps 6e9 mm $3 polyps with diameter <10 mm Adenomas $6 mm Adenomas $10 mm Adenomas 6e9 mm $3 adenomas with diameter <10 mm Per-patient with positive FOBT (n¼50) Overall Polyps $6 mm Polyps $10 mm Polyps 6e9 mm $3 polyps with diameter <10 mm Adenomas $6 mm Adenomas $10 mm Adenomas 6e9 mm $3 adenomas with diameter <10 mm

Sensitivity

Specificity

PPV

NPV

52 64 80 55 11 68 78 62 17

(42 to 62) (48 to 78) (56 to 94) (36 to 74) (1 to 35) (49 to 83) (52 to 94) (38 to 82) (2 to 48)

84 94 99 95 97 94 98 95 98

(77 (90 (96 (91 (95 (90 (96 (92 (95

to to to to to to to to to

89) 97) 100) 98) 99) 97) 100) 98) 99)

68 70 84 59 25 64 78 54 29

(56 to 78) (53 to 83) (60 to 97)* (39 to 78) (3 to 65) (46 to 79) (52 to 94)* (33 to 74) (4 to 71)

73 92 98 94 93 95 98 97 96

(66 (88 (96 (90 (90 (91 (96 (93 (93

to to to to to to to to to

79)** 96) 100) 97) 96) 97) 100) 98) 98)

54 69 73 65 14 68 75 62 20

(47 to 61) (58 to 78) (56 to 86) (51 to 77) (7 to 24) (56 to 79) (57 to 89) (46 to 76) (8 to 39)

79 90 95 94 99 90 95 94 98

(72 (86 (93 (91 (97 (86 (93 (91 (96

to to to to to to to to to

84) 93) 97) 96) 100) 92) 97) 96) 99)

75 62 60 63 67 55 57 53 46

(68 (52 (44 (49 (38 (44 (41 (38 (19

59 92 97 95 85 94 98 96 94

(53 (89 (95 (92 (81 (91 (96 (93 (92

to to to to to to to to to

65)** 95) 99) 97) 88) 96) 99) 98) 96)

74 88 92 75 25 88 92 75 25

(56 to 87) (62 to 98) (64 to 100) (19 to 99) (3 to 65) (62 to 98) (64 to 100) (19 to 99) (1 to 81)

75 91 97 96 98 91 97 96 98

(48 (76 (86 (85 (87 (76 (86 (85 (88

to to to to to to to to to

93) 98) 100) 99) 100) 98) 100) 99) 100)

86 82 92 60 67 82 92 60 50

(69 to 96) (57 to 96) (64 to 100)* (15 to 95) (9 to 99) (57 to 96) (64 to 100)* (15 to 95) (1 to 99)

57 94 97 98 87 94 97 98 94

(34 (80 (86 (88 (74 (80 (86 (88 (83

to to to to to to to to to

78)** 99) 100) 100) 95) 99) 100) 100) 99)

to to to to to to to to to

81) 72) 74)* 75) 88) 66) 72)* 67) 75)

Values are % (95% CI). *p<0.05; **p<0.01 (c2 test and Fisher exact test). FOBT, faecal occult blood test; NPV, negative predictive value; PPV, positive predictive value.

Our model accounted for 47% of the variance of per-radiologist diagnostic sensitivity at CTC for polyps $6 mm, on a per-lesion analysis. The detection rate achieved by radiologists for identiﬁcation of polyps $6 mm during the training session was the only predictor that was slightly signiﬁcant (p<0.046). Correlation data between the detection rates obtained during the training set and for the cases included in the daily practice are available in appendix 4 (online only).

Miscellaneous data Accuracy of videocolonoscopy compared with CTC The segmental unblinded colonoscopy method allowed us to compare the accuracy of videocolonoscopy alone with that of CTC alone. On a per-patient analysis, for patients with polyps of any size, the accuracy of videocolonoscopy alone was 96.2% (95% CI 94.6% to 97.5%), and for CTC alone it was 68.0% (95% CI 64.5% to 71.3%). For patients with polyps $6 mm, these values were 99.5% (95% CI 98.6% to 99.9%) vs 87.0 (95% CI 84.3 to 89.3), and for patients with polyps $10 mm, the values were 99.7% (95% CI 99.0% to 99.9%) vs 95.0 (95% CI 93.2 to 96.4).

Effect of assessed quality of colon preparation Colonic preparation was described by the endoscopist as excellent or good in 63% of cases, fair in 28%, and poor or inadequate in 8.5% and 0.5% of patients. Only 23 procedures (3.1%) were classiﬁed as C0 in the C-RAD classiﬁcation, where C0 includes inadequate preparation or insufﬂation for satisfactory 662

interpretation of CTC results; in 20 of our cases it was due to inadequate insufﬂation. So, these 23 patients were described at optical colonoscopy as having excellent or good preparation in 65% of cases, fair in 30%, and poor or inadequate in 5%. The per-patient sensitivity and PPV for detecting patients harbouring polyps $6 mm were 69% (95% CI 60% to 77%) and 66% (95% CI 57% to 74%) at CTC for those assessed at optical colonoscopy as having excellent, good or fair colonic preparation, versus 78% (95% CI 49% to 95%) and 79% (95% CI 49% to 95%) for those with poor or inadequate preparation, respectively.

Estimates of polyp size Estimates of polyp diameter from optical colonoscopy and CTC correlated signiﬁcantly, and results are available online in appendix 7.

Polyp shape: effect on CTC accuracy

Diagnostic value results for ﬂat versus non-ﬂat (pedunculated or sessile) lesions (excluding stenosing or massive tumorous lesions that could not be categorised according to the Paris classiﬁcation) are available in appendix 2 online. There were no signiﬁcant differences in sensitivity for polyp shape.

DISCUSSION Our ﬁndings on per-lesion or per-patient analyses are similar to those of two meta-analyses on CTC1 2: comparisons may be made more easily with that of Rosman and Korsten1 or Mulhall et al2 Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon Table 4 Sensitivity (per-patient) of radiologists in detection of lesions of different types and sizes by computed tomographic colonoscopy (CTC), according to performance at initial training and subsequent experience Sensitivity of radiologists in upper half of distribution

Sensitivity of radiologists in lower half of distribution Variable of interest

Lesion group

%

95% CI

%

95% CI

Training performance: % detection of polyps of any size in training set (median was 49%)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

52 62 68 61* 68

44 50 48 47 47

to to to to to

60 74 84 74 85

58 76 86 80* 87

51 65 71 68 72

to to to to to

65 85 95 89 96

Training performance: % detection of polyps $6 mm in training set (median was 61%)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

51 63 69 60* 68

43 50 49 46 47

to to to to to

59 74 85 74 85

58 75 85 79* 87

52 64 71 67 72

to to to to to

65 84 94 88 96

Case volume in the study (median number was 18)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

55 56 50 55 25*

40 to 69 31 to 78 12 to 88 23 to 83 1 to 81

55 71 81 72 83*

50 63 70 63 71

to to to to to

61 79 90 81 92

Potential number of polyps for detection (median number was 19)

Overall Polyps $6 mm Polyps $10 mm Adenomas $6 mm Adenomas $10 mm

58 68 56 70 57

43 47 21 46 18

55 70 82 71 82

49 61 70 61 70

to to to to to

61 78 91 80 91

to to to to to

72 85 86 88 90

The radiologists were divided into two subgroups, according to the median of the distribution of the variable of interest, and the pooled sensitivity was found for each subgroup. *p<0.03 (c2 test and Fisher exact test).

who found sensitivities for polyps >5 mm or >10 mm, respectively, of 0.73 (95% CI 0.62 to 0.85) and 0.81 (95% CI 0.74 to 0.87). Our results were obtained with CTC performed after standard large-bowel preparation,14 15 although CTC data with less purgation have been rarely described.7 17 18 The recent joint guideline from the US Multi-Society Task Force on CRC emphasises that clinicians should make patients aware of the full range of screening options and at least offer the choice between a screening test that is primarily effective at early cancer detection and one that is effective at both early detection and cancer prevention, including CTC19 20 in a similar sensitivty than videocolonoscopy.21e23 Nevertheless, despite this endorsement, the Centers for Medicare and Medicaid Services (CMS) recently decided to deny coverage of CTC for CRC screening, partly because of the heterogeneity of the multicentre results and the absence of clear deﬁnitions of expertise and certiﬁcation.24 Although most studies deﬁned the expertise of the participating radiologists, the deﬁnition related mostly to previous experience as speciﬁed by a minimal number of CTCs analysed before participation in the trial. This number varied from 1000,25 25e300 26 27 38 or 50 12 28 for the largest studies. Some national guidelines 29 30 recommend that, for certiﬁcation, the radiologist should analyse 50e75 CTCs and then take a test based on a set of 20 cases, whereas the American Gastroenterological Association recommends training that involves review and interpretation of at least 75 CTCs with endoscopic correlation, and further formal mentoring with at least 25e50 additional cases.3 In fact, all these guidelines rely on either expert assessment (with expert deﬁned as a radiologist who had interpreted at least 500 CTC examinations12) and had achieved a detection rate higher than Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

90% for polyps measuring $10 mm in diameter. In the ACRIN study,16 of the 20 radiologists who met the initial criteria, the 15 with the highest score on the qualifying examination were subsequently invited to participate in the study. In fact, although several studies31e33 have described the impact of expertise on CTC accuracy, only four12 15 34 35 have closely analysed the impact of initial training on polyp detection rate. Some studies have reported on the increase in accuracy after training with 25 cases,31 or on the accuracy of radiologists after training sets of 50 cases 26 or 75 cases. 32 Other studies have focused on the description of the learning curve, comparing radiologists with non-radiologists 36 or expert versus non-expert radiologists,26 or making both comparisons,29 between experienced radiologists and novice radiologists and technologists who had undergone training on 50 cases. Another study compared the results for an experienced reader with two inexperienced readers using different modes of analysis.34 However, in daily practice, previous experience with CTC should have an effect: the sensitivity for polyp detection was higher (53e93%) for radiologists with greater previous experience (5e320 CTCs).4 37 Interestingly, our ﬁndings are in accordance with those of Johnson et al,16 although their study focuses on the sensitivity estimates for each of the 15 radiologists and detection rate for polyps >1 cm. However, the study did not determine whether sensitivity was related to experience or to the scores obtained during the test examinations, because they selected only radiologists with a high sensitivity for polyps >1 cm. In fact, one study had determined the accuracy of CTC reading before and after a training set38; in this trial the rate of detection (whatever the size of the lesions) post- minus pre-training signiﬁcantly increases by 18%, whatever the expertise of the radiologist. 663

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon We cannot compare our data directly, but there is an increase in accuracy during the inclusion especially when we compared sensitivity of the training set with cases included in our study by each radiologist. Finally, the questions about CTC screening included: should CTC be read by an expert radiologist only or by either a trained radiologist or gastroenterologist? This question has been addressed in a recent study,39 which found no statistically signiﬁcant difference in sensitivity of speciﬁcity between gastroenterologists (and radiologists) for detection of polyps whatever their diameter. Our study included participants with different risk levels for CRC: increased risk (ie, with familial or personal history of CRC or adenoma); average risk (ie, with no history of colorectal neoplasia) which included patients with digestive symptoms and asymptomatic participants who wanted or needed a colonic examination; and patients with a recent positive FOBT. Our univariate analysis did demonstrate a greater sensitivity and PPV for the last subgroup; these values were higher only in relation to lesions $10 mm. All of the radiologists in our trial were required to have trained on an initial set of 52 CTC cases, to obtain the minimal level of competence. The cases were speciﬁcally chosen to teach identiﬁcation of lesions that were difﬁcult to detect; thus the median detection rates in this set of 52 cases were 49%, 61% and 65% for polyps of any size, for those $6 mm, and for those $10 mm, respectively. Comparatively few studies have determined the adequate level of training for technical competence in screening or diagnostic optical colonoscopy. A recent multicentre prospective evaluation of the learning curve concluded that technical competence in colonoscopy required experience with at least 150 cases for a successful caecal intubation rate higher than 90%.30 The mean differences in estimated diameter were higher (644%) for polyps 6e9 mm and those $10 mm (range from e48% to +22%). It has previously been reported that the average absolute difference between CTC and open biopsy forceps measurements was 2 mm.9 Thus, the calculated accuracy of CTC in our study for polyps >6 mm should not be inﬂuenced greatly by the CTC method used to determine polyp diameter. To conclude, the less invasive nature of CTC may be attractive to patients, and the use of this modality may improve screening compliance rates because many people are reluctant to undergo colonoscopy because of its perceived inconvenience or discomfort, or because of embarrassment.40 The most relevant ﬁnding of our study, reported according to the Standard for Reporting of Diagnostic Accuracy (online appendix 8),41 was the value of the radiologist’s polyp detection rate with the training set in predicting accuracy at CTC for detecting polyps $6 mm, independently of the radiologist’s case volume or the number of polyps harboured in these cases.

11

Department of Radiology, University Hospital, CHU, Bordeaux Saint-Andre´, France Department of Gastroenterology, University Hospital, CHU Angers, France 13 Department of Radiology, Assistance Publique-Hoˆpitaux de Paris, Beaujon, France 14 Department of Gastroenterology, Assistance Publique-Hoˆpitaux de Paris, Bichat, France 15 Department of Gastroenterology, Assistance Publique-Hoˆpitaux de Paris, Henri Mondor, France 16 Department of Gastroenterology, University Hospital, CHU Reims, France 17 Department of Radiology, University Hospital, CHU Limoges, France 18 Department of Gastroenterology, University Hospital, CHU Montpellier, France 19 Department of Radiology, University Hospital, CHU Lyon Edouard Herriot, France 20 Department of Gastroenterology, Assistance Publique-Hopitaux de Paris, Lariboisiere, Paris 21 Department of Radiology, Hospital CH, Pontivy, France 22 Department of Gastroenterology, University Hospital, CHU Bordeaux Saint-Andre, France 23 Department of Radiology, University Hospital, CHU Angers, France 24 Department of Gastroenterology, Assistance Publique-Hopitaux de Paris, Beaujon, France 25 Department of Radiology, Assistance Publique-Hopitaux de Paris, Bichat, France 26 Department of Radiology, University Hospital, CHU de Bordeaux Haut-Le´veˆque, France 27 Department of Radiology, Assistance Publique-Hoˆpitaux de Paris, St Louis, France 28 Department of Radiology, University Hospital, CHU 1 Rennes, France 12

Author footnote: The following also participated in this study: Anne Ganivet (University Hospital, CHU Rennes), Franck Pilleul (University Hospital, CHU Lyon Edouard Herriot), Se´verine Poupeney (Assistance Publique-Hoˆpitaux de Paris (APHP), Lariboisie`re), Pascal Burtin (University Hospital, CHU Angers), Vincent Barrau (Assistance Publique-Hoˆpitaux de Paris, APHP Beaujon), Thierry Vallot (Assistance Publique-Hoˆpitaux de Paris, APHP Bichat), Cle´ment Subtil (Centre Hospitalier Universitaire (CHU) de Bordeaux Haut-Le´veˆque), Marion Simon (Assistance Publique-Hoˆpitaux de Paris, APHP Saint-Louis), Ce´dric De Bazelaire (Assistance Publique-Hoˆpitaux de Paris, APHP Saint-Louis), Jean Marc Gornet (Assistance Publique-Hoˆpitaux de Paris, APHP Saint-Louis), Nicolas Sellier (Assistance Publique-Hoˆpitaux de Paris, APHP Jean Verdier), Michel Beaugrand (Assistance Publique-Hoˆpitaux de Paris, APHP Jean Verdier), Sophie Maitre (Assistance Publique-Hoˆpitaux de Paris, APHP Antoine Be´cle`re), Amani Asnacios (Assistance Publique-Hoˆpitaux de Paris, APHP Antoine Be´cle`re), Bertrand Bessoud (Assistance Publique-Hoˆpitaux de Paris, APHP Biceˆtre), Anne-Sophie Rangheard (Assistance Publique-Hoˆpitaux de Paris, APHP Biceˆtre), Isabelle Boytchev (Assistance Publique-Hoˆpitaux de Paris, APHP Biceˆtre), Nathalie Siauve (Assistance Publique-Hoˆpitaux de Paris, APHP HEGP), Francis Bloch (Assistance Publique-Hoˆpitaux de Paris, APHP HEGP), Ache`ne Belkacem (Assistance Publique-Hoˆpitaux de Paris, APHP Louis Mourier), Benoit Coffin (Assistance Publique-Hoˆpitaux de Paris, APHP Louis Mourier), Elsa Guillot (University Hospital CHU Lyon Sud), Jean-Christophe Saurin (University Hospital CHU Lyon Sud), Philippe Beaurain (Private Hospital Marseille Saint-Joseph), Christian Boustie`re (Private Hospital Marseille Saint-Joseph), Denis Marion (University Hospital CHU Lyon Hotel Dieu), Franc¸ois Bailly (University Hospital CHU Lyon Hotel Dieu), Vale´rie Laurent (University Hospital CHU Nancy), Ge´rard Gay (University Hospital CHU Nancy), Ce´line Savoye-Collet (University Hospital CHU Rouen), Emmanuel Ben Soussan (University Hospital CHU Rouen), Franck Boudghe`ne (Assistance Publique-Hoˆpitaux de Paris, APHP Tenon), Jean-Didier Grange´ (Assistance Publique-Hoˆpitaux de Paris, APHP Tenon), Medhi Cadi (Assistance Publique-Hoˆpitaux de Paris, APHP Pitie´ Salpeˆtrie`re), Robert Chollet (Assistance Publique-Hoˆpitaux de Paris, APHP Pitie´ Salpeˆtrie`re). Funding This study was funded by a grant from the National Cancer Institute of France (INCa (Institut National du Cancer)), under the French Department of Health STIC (2005) programme, the French Society of Digestive Endoscopy (SFED (Socie´te´ Franc¸aise d’Endoscopie Digestive)), and SFR (Socie´te´ Franc¸aise de Radiologie). The STIC programme supports research into new high-cost techniques (‘Soutien aux Techniques Innovantes et Couˆteuses’). Competing interests None. Patient consent Obtained.

Author affiliations: 1 Department of Gastroenterology, University Hospital, CHU Rennes, France 2 UMR CNRS 6211, CREM, Faculte´ des Sciences Economiques de Rennes, Rennes, France 3 Department of Radiology, Assistance Publique-Hoˆpitaux de Paris, Henri Mondor, France 4 Department of Public Health and Biostatistic, University Hospital, CHU Rennes, France 5 Department of Radiology, University Hospital, CHU Reims, France 6 Department of Gastroenterology, University Hospital, CHU Limoges, France 7 Department of Radiology, University Hospital, CHU Montpellier, France 8 Department of Gastroenterology, University Hospital, CHU Lyon Edouard Herriot, France 9 Department of Radiology, Assistance Publique-Hoˆpitaux de Paris, Lariboisie`re, France 10 Department of Gastroenterology, Hospital CH, Pontivy, France 664

Ethics approval This study was registered with the Health General Directory (No DGS: 2006/0412; 11 August 2006) and conducted with the approval of the Ethics and Subjects Protection Committee (CPP) (6 July 2006). Contributors DH and YG designed and coordinated the project, DH wrote the paper, and FR and SH conducted the statistical analysis. The remaining authors contributed data for most of the patients included in the study. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1.

Rosman AS, Korsten MA. Meta-analysis comparing CT colonography, air contrast barium enema, and colonoscopy. Am J Med 2007;120:203e10.

Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Colon 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20.

Mulhall BP, Veerappan GR, Jackson JL. Meta-analysis: computed tomographic colonography. Ann Intern Med 2005;142:635e50. Rockey DC, Barish M, Brill JV, et al. Standards for gastroenterologists for performing and interpreting diagnostic computed tomographic colonography. Gastroenterology 2007;133:1005e24. Burling D, Halligan S, Atchley J, et al. Colonography: interpretative performance in a non-academic environment. Clin Radiol 2007;62:424e31. Cotton PB, Durkalski VL, Pineau BC, et al. Computed tomographic colonography (virtual colonoscopy): a multicenter comparison with standard colonoscopy for detection of colorectal neoplasia. JAMA 2004;291:1713e19. Dachman AH, Kelly KB, Zintsmaster MP, et al. Formative evaluation of standardized training for CT colonographic image interpretation by novice readers. Radiology 2008;249:167e77. Mahgerefteh S, Fraifeld S, Blachar A, et al. CT colonography with decreased purgation: balancing preparation, performance, and patient acceptance. AJR Am J Roentgenol 2009;193:1531e9. Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 2003;349:2191e200. Gupta S, Durkalski V, Cotton P, et al. Variation of agreement in polyp size measurement between computed tomographic colonography and pathology assessment: clinical implications. Clin Gastroenterol Hepatol 2008;6:220e7. Anonymous. The Paris endoscopic classification of superficial neoplastic lesion. Gastrointest Endosc 2003;58:S3e43. Dixon MF. Gastrointestinal epithelial neoplasia: Vienna revisited. Gut 2002;51:130e1. Zalis ME, Barish MA, Choi JR, et al. Working Group on Virtual Colonoscopy. CT colonography reporting and data system: a consensus proposal. Radiology 2005;236:3e9. Aronchick CA, Lipshutz WH, Wright SH, et al. A novel tableted purgative for colonoscopic preparation: efficacy and safety comparisons with Colyte and Fleet Phospho-Soda. Gastrointest Endosc 2000;52:346e52. Kim DH, Pickhardt PJ, Taylor AJ, et al. CT colonography versus colonoscopy for the detection of advanced neoplasia. N Engl J Med 2007;357:1403e12. Rockey DC, Paulson E, Niedzwiecki D, et al. Analysis of air contrast barium enema, computed tomographic colonography, and colonoscopy: prospective comparison. Lancet 2005;365:305e11. Johnson CD, Chen MH, Toledano AY, et al. Accuracy of CT colonography for detection of large adenomas and cancers. N Engl J Med 2008;359:1207e17. Taylor SA, Slater A, Burling DN, et al. CT colonography: optimisation, diagnostic performance and patient acceptability of reduced-laxative regimens using bariumbased faecal tagging. Eur Radiol 2008;18:32e42. Jensch S, de Vries AH, Peringa J, et al. CT colonography with limited bowel preparation: performance characteristics in an increased-risk population. Radiology 2008;247:122e32. Levin B, Lieberman DA, McFarland B, et al; American Cancer Society Colorectal Cancer Advisory Group; US Multi-Society Task Force; American College of Radiology Colon Cancer Committee. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Gastroenterology 2008;134:1570e95. Heresbach D, Manfredi S, d’Halluin PN, et al. Review in depth and meta-analysis of controlled trials on colorectal cancer screening by faecal occult blood test. Eur J Gastroenterol Hepatol 2006;18:427e33.

Gut 2011;60:658e665. doi:10.1136/gut.2010.225623

21. 22. 23. 24. 25. 26. 27. 28. 29.

30. 31. 32. 33. 34.

35. 36. 37. 38. 39. 40. 41.

Heresbach D, Barrioz T, Lapalus MG, et al. Miss rate for colorectal neoplastic polyps: a prospective multicenter study of back-to-back video colonoscopies. Endoscopy 2008;40:284e90. Imperiale TF, Glowinski EA, Juliar BE, et al. Variation in polyp detection rates at screening colonoscopy. Gastrointest Endosc 2009;69:1288e95. van Rijn JC, Reitsma JB, Stoker J, et al. Polyp miss rate determined by tandem colonoscopy: a systematic review. Am J Gastroenterol 2006;101:343e50. Garg S, Ahnen DJ. Is computed tomographic colonography being held to a higher standard? Ann Intern Med 2010;152:178e81. Johnson CD, Fletcher JG, MacCarty RL, et al. Effect of slice thickness and primary 2D versus 3D virtual dissection on colorectal lesion detection at CT colonography in 452 asymptomatic adults. AJR Am J Roentgenol 2007;189:672e80. Taylor SA, Halligan S, Burling D, et al. CT colonography: effect of experience and training on reader performance. Eur Radiol 2004;14:1025e33. Graser A, Stieber P, Nagel D, et al. Comparison of CT colonography, colonoscopy, sigmoidoscopy, and fecal occult blood tests for the detection of advanced adenoma in an average risk population. Gut 2009;58:241e8. Van Gelder RE, Venema HW, Serlie IW, et al. CT colonography at different radiation dose levels: feasibility of dose reduction. Radiology 2002;224:25e33. European Society of Gastrointestinal and Abdominal Radiology CT Colonography Study Group Investigators. Effect of directed training on reader performance for CT colonography: multicenter study. Radiology 2007;242:152e61. Lee SH, Chung IK, Kim SJ, et al. An adequate level of training for technical competence in screening and diagnostic colonoscopy: a prospective multicenter evaluation of the learning curve. Gastrointest Endosc 2008;67:683e9. Gluecker T, Meuwly JY, Pescatore P, et al. Effect of investigator experience in CT colonography. Eur Radiol 2002;12:1405e9. Spinzi G, Belloni G, Martegani A, et al. Computed tomographic colonography and conventional colonoscopy for colon diseases: a prospective, blinded study. Am J Gastroenterol 2001;96:394e400. Slater A, Taylor SA, Tam E, et al. Reader error during CT colonography: causes and implications for training. Eur Radiol 2006;16:2275e83. Fisichella VA, Ja¨derling F, Horvath S, et al. Primary three-dimensional analysis with perspective-filet view versus primary two-dimensional analysis: evaluation of lesion detection by inexperienced readers at computed tomographic colonography in symptomatic patients. Acta Radiol 2009;50:244e55. Jensch S, van Gelder RE, Florie J, et al. Performance of radiographers in the evaluation of CT colonographic images. AJR Am J Roentgenol 2007;188:249e55. Bodily KD, Fletcher JG, Engelby T, et al. Nonradiologists as second readers for intraluminal findings at CT colonography. Acad Radiol 2005;12:67e73. Burling D, Halligan S, Altman DG, et al. CT colonography interpretation times: effect of reader experience, fatigue, and scan findings in a multi-centre setting. Eur Radiol 2006;16:1745e9. Taylor SA, Burling D, Roddie M, et al. Computer-aided detection for CT colonography: incremental benefit of observer training. Br J Radiol 2008;81:180e6. Young PE, Ray QP, Hwang I, et al. Gastroenterologists’ interpretation of CTC: a pilot study demonstrating feasibility and similar accuracy compared with radiologists’ interpretation. Am J Gastroenterol 2009;104:2926e31. McFarland EG, Fletcher JG, Pickhardt P, et al. American College of Radiology. ACR Colon Cancer Committee white paper: status of CT colonography 2009. J Am Coll Radiol 2009;6:756e72. Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. Standards for Reporting of Diagnostic Accuracy. BMJ 2003;326:41e4.

665

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas

Endoscopic retrograde pancreatography criteria to diagnose autoimmune pancreatitis: an international multicentre study Aravind Sugumar,1 Michael J Levy,1 Terumi Kamisawa,5 George J M Webster,6 Myung-Hwan Kim,3 Felicity Enders,1 Zahir Amin,7 Todd H Baron,1 Mike H Chapman,6 Nicholas I Church,6 Jonathan E Clain,1 Naoto Egawa,5 Gavin J Johnson,6 Kazuichi Okazaki,4 Randall K Pearson,1 Stephen P Pereira,6 Bret T Petersen,1 Samantha Read,7 Raghuwansh P Sah,1 Neomal S Sandanayake,6 Naoki Takahashi,2 Mark D Topazian,1 Kazushige Uchida,4 Santhi Swaroop Vege,1 Suresh T Chari1 See Commentary, p 565 < Additional tables and figure

are published online only. To view these files please visit the journal online (http://gut.bmj. com). 1

Division of Gastroenterology and Hepatology, Mayo Clinic College of Medicine, Rochester, MN, USA 2 Division of Radiology, Mayo Clinic College of Medicine, Rochester, MN, USA 3 University of Ulsan College of Medicine, Asan Medical Center, Seoul, South Korea 4 Department of Gastroenterology and Hepatology, Kansai Medical University, Hirakata Hospital, Osaka, Japan 5 Department of Internal Medicine, Tokyo Metropolitan Komagome Hospital, Tokyo, Japan 6 Department of Gastroenterology, University College Hospital, London, UK 7 Department of Radiology, University College Hospital, London, UK Correspondence to Suresh T Chari, Mayo Clinic College of Medicine, Division of Gastroenterology and Hepatology, 200 First St. SW, Rochester, MN 55905, USA; [email protected] Revised 31 October 2010 Accepted 1 November 2010 Published Online First 3 December 2010

666

ABSTRACT Background Characteristic pancreatic duct changes on endoscopic retrograde pancreatography (ERP) have been described in autoimmune pancreatitis (AIP). The performance characteristics of ERP to diagnose AIP were determined. Methods The study was done in two phases. In phase I, 21 physicians from four centres in Asia, Europe and the USA, unaware of the clinical data or diagnoses, reviewed 40 preselected ERPs of patients with AIP (n¼20), chronic pancreatitis (n¼10) and pancreatic cancer (n¼10). Physicians noted the presence or absence of key pancreatographic features and ranked the diagnostic possibilities. For phase II, a teaching module was created based on features found most useful in the diagnosis of AIP by the four best performing physicians in phase I. After a washout period of 3 months, all physicians reviewed the teaching module and reanalysed the same set of ERPs, unaware of their performance in phase I. Results In phase I the sensitivity, specificity and interobserver agreement of ERP alone to diagnose AIP were 44, 92 and 0.23, respectively. The four key features of AIP identified in phase I were (i) long (>1/3 the length of the pancreatic duct) stricture; (ii) lack of upstream dilatation from the stricture (<5 mm); (iii) multiple strictures; and (iv) side branches arising from a strictured segment. In phase II the sensitivity (71%) of ERP significantly improved (p<0.05) without a significant decline in specificity (83%) (p>0.05); the interobserver agreement was fair (0.40). Conclusions The ability to diagnose AIP based on ERP features alone is limited but can be improved with knowledge of some key features.

INTRODUCTION Autoimmune pancreatitis (AIP) is a rare but increasingly recognised form of chronic pancreatitis that predominantly affects men in their ﬁfth and sixth decades.1 Since it often presents with obstructive jaundice and pancreatic enlargement, the chief differential diagnosis of AIP is pancreatic cancer (PaC).1e3 As AIP is a relatively uncommon disease, its diagnosis in patients with suspected PaC can be quite challenging.2 3 In addition, distinguishing AIP from usual chronic pancreatitis (CP; both clinically and radiologically) is also difﬁcult.2

Significance of this study What is already known about this subject?

< Autoimmune pancreatitis (AIP) is a distinct form

of chronic pancreatitis.

< AIP most often mimics pancreatic cancer in its

presentation.

< Multiple diagnostic criteria exist for AIP.

What are the new findings? are specific endoscopic retrograde pancreatography (ERP) features in AIP that are helpful in the diagnosis. < The ability to diagnose AIP based on ERP features alone is limited but can be improved with knowledge of some key features. < We have proposed an algorithm to diagnose AIP as well as to differentiate it from pancreatic cancer. < There

How might it impact on clinical practice in the foreseeable future? < This is the first multicentre, international study addressing the role of ERP in AIP. Thus this study will be one of the building blocks while developing international consensus criteria for AIP.

Since Yoshida et al coined the term ‘autoimmune pancreatitis’ in 1995, the diagnostic value of imaging features of AIP has been recognised.4 The characteristic pancreatic imaging features include parenchymal features seen on cross-sectional imaging and ductal features seen on endoscopic retrograde pancreatography (ERP). In fact when the Japan Pancreas Society (JPS) proposed diagnostic criteria for AIP, it considered the presence of both parenchymal and ERP features mandatory for diagnosis of AIP.5 More recently, the HISORt criteria for diagnosis of AIP were proposed by clinicians at the Mayo Clinic, Rochester, USA. While recognising the value of imaging features, the HISORt criteria do not mandate their presence for diagnosis of AIP.6 7 Investigators worldwide have come together to form an Autoimmune Pancreatitis International Gut 2011;60:666e670. doi:10.1136/gut.2010.207951

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas Figure 1 Overview of the methods. AIP, autoimmune pancreatitis; CP, chronic pancreatitis; ERP, endoscopic retrograde pancreatography; PaC, pancreatic cancer.

164 ERP’s screened for quality by 1 endoscopist unaware of diagnosis

48 good quality ERPs selected (20 AIP, 10 CP, 10 PaC and 8 duplicates)

Randomized and shuffled

Phase II

Phase I

Sent out to 4 centers (US, Japan, Korea and UK)

21 physicians (8 US, 8 UK, 5 Asia)

3 month washout Period Sent out to 4 centers Phase I results + Input from top performing. Physicians = Teaching Module

21 physicians

Data analyzed

Data analyzed

Cooperative Study group (APICS) to better study this rare disease. Since there is substantial disparity in the use of ERP to diagnose AIP between the Asian and US criteria, one of the ﬁrst goals of the APICS was to evaluate the performance characteristics (sensitivity, speciﬁcity and interobserver agreement) of ERP for diagnosis of AIP under the auspices of an international multicentre trial. Here we report the results of the ﬁrst study proposed by this group.

MATERIALS AND METHODS The study was approved by the Mayo Clinic’s institutional review board. The study was done in two phases. A total of 21 physicians from four centres in Asia (Japan and Korea), the UK and the USA participated in interpretation of ERP images (readers) in both phases of the study. The lead and senior authors (AS and STC) did not interpret images. A blinded randomised trial design was used for both phases (ﬁgure 1). The diagnosis of AIP was made using HISORt and Japanese criteria, depending on the centre where the patient was diagnosed (Supplementary table 4).5 7 Approximately 35% of patients did not have immunoglobulin G4 (IgG4) levels recorded. Many of these patient were diagnosed with AIP before IgG4 was found to be associated with AIP, but all these patients had the usual clinical course of AIP. The patients with PaC had a conﬁrmed tissue diagnosis and/ or the typical clinical course. The diagnosis of CP was made with a combination of the clinical presentation with pancreatitis and characteristic ERP ﬁndings (Cambridge III or IV).8 9

Phase I ERPs (n¼164) from histologically and/or clinically conﬁrmed cases of AIP, CP and PaC were solicited from three participating

Table 1

centres (the USA, Japan and the UK).The ERP images were screened for quality, including brightness, adequate contrast enhancement and adequacy of duct ﬁlling by an experienced endoscopist (MJL) unaware of clinical diagnoses. A ﬁnal set of 48 pancreatograms (20 AIP, 10 CP, 10 PaC and 8 duplicates) were selected by the senior author (STC). We ensured that all three centres had equal representation in the ﬁnal set of images selected for interpretation. To ascertain intraobserver variability, eight duplicate images (4 AIP and 2 each of PaC and CP) were included in the set of 48 pancreatograms. A random number generator was used to generate numbers for the 48 images. These images were then shufﬂed. The ﬁnal set of randomised and shufﬂed images was sent to 21 physicians in the USA, the UK, Japan and South Korea on a compact disc. These readers were not privy to clinical data or the underlying diagnoses of the images sent. Each physician viewed the ﬁnal set of images on a computer. Most of them completed the interpretation in one sitting and ﬁlled out a data abstraction sheet (Supplementary table 1) for each image. This sheet enumerated ERP features of CP, AIP and PaC which have been previously described.3 10 11 The physicians were asked to provide the most probable diagnoses based on ERP ﬁndings. Up to three diagnoses could be listed by percentage conﬁdence (Supplementary table 1). The sum of the conﬁdences had to add up to a 100%. Thus acceptable percentage conﬁdence combinations were 95%e5%, 75%e25%, 50%e50%, and 50%e 25%e25%. To be considered a correct interpretation, a given pancreatogram had to be read with at least 75% conﬁdence for that condition. The data abstraction sheets were then collated to compute the performance characteristics of ERP for diagnosis of AIP (table 1).

Phase I results

Sensitivity (95% CI) Specificity (95% CI) Interobserver agreement (k)

Overall

USA (N[8)

UK (N[8)

Asia (N[5)

Asia vs rest

44% (40% to 49%) 92% (89% to 95%) 0.23

33% (26% to 41%) 92% (86% to 96%) 0.16

40% (32% to 48%) 95% (90% to 98%) 0.16

71% (61% to 79%) 89% (81% to 94%) 0.52

p<0.05 NS

NS, non-significant.

Gut 2011;60:666e670. doi:10.1136/gut.2010.207951

667

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas

Figure 2 Key features of autoimmune pancreatitis in a pancreatogram. Feature 1: long (>1/3 the length of the pancreatic duct) narrow stricture. Feature 2: lack of upstream dilatation from the stricture (<5 mm).

Figure 3 Key features of autoimmune pancreatitis in a pancreatogram. Feature 3: multiple strictures. Feature 4: side branches arising from the stricture site.

Creation of a teaching module

reference to the external gold standard (Asian and HISORt criteria) and the k statistic was computed to determine the agreement between physicians from each centre. A k statistic for intraobserver agreement was computed in phase I alone. A p value <0.05 was considered statistically signiﬁcant. In addition, in phase II, the sensitivity and speciﬁcity of the four ERP features in AIP in differing combinations was computed (table 4).

From the phase I data it was apparent that the overall sensitivity and speciﬁcity for the entire group (N¼21) was poor, but there were four physicians (not all from Asia) whose interpretations had a high sensitivity and speciﬁcity for diagnosing AIP based solely on an ERP. These physicians were asked to identify key ERP features which aided them in correctly diagnosing AIP. This led to the identiﬁcation of four key ERP features (ﬁgures 2 and 3). A Power Point teaching module was then created highlighting these observations and the overall (not individual) phase I results. The teaching module did not include any images that were used in phase I

Phase II After a 3 month washout period, we requested all readers in phase I to interpret the same 40 pancreatograms used in phase I (20 AIP, 10 CP and 10 PaC) without the eight duplicate images. A random number generator was again used to randomise the 40 images. These images were then shufﬂed. The ﬁnal set of randomised and shufﬂed images was sent to 21 physicians in the USA, the UK, Japan and South Korea on a compact disc. All 21 physicians were asked to review the teaching module before reanalysing the pancreatograms. Apart from the information requested from the readers in phase I, they were speciﬁcally asked to comment on the presence or absence of the four key ERP features identiﬁed in phase I. In addition to computing the overall performance characteristics of ERP to diagnose AIP (table 2), we also calculated the performance characteristics of the four key ERP features in diagnosing AIP (table 3).

RESULTS Phase I The overall sensitivity, speciﬁcity and interobserver agreement of ERP to diagnose AIP were 44, 92 and 0.23, respectively, across all participating centres (table 1). There was wide variation in the sensitivity of the interpreters to diagnose AIP based on ERP features across the four participating centres. The sensitivity among Asian physicians was signiﬁcantly higher than among non-Asian physicians. However, the difference in speciﬁcity across the centres was not signiﬁcant. In addition to computing the interobserver variability, we also computed the intraobserver variability (from the eight duplicate ﬁlms), but this was not statistically signiﬁcant. The stratiﬁed results based on the specialty of the reader are shown in Supplementary table 2. Of the 21 reviewers, we identiﬁed four top performing physicians who had high sensitivities and speciﬁcities (Supplementary table 2). In the opinion of the top performing physicians the following four ERP features were most helpful to diagnose AIP (ﬁgures 2, 3 and table 3) (i) long (>1/3 the length of the pancreatic duct) stricture; (ii) lack of upstream dilatation from the stricture (<5 mm); (iii) multiple strictures; and (iv) side branches arising from a strictured segment.

Statistical methods We used JMP Version 8 for computing the sensitivity, speciﬁcity and k statistic for interobserver agreement in both phase I and phase II. A 232 table was constructed for each reader in both phases for this purpose. The sensitivity and speciﬁcity are with Table 2

Phase II The overall sensitivity, speciﬁcity and interobserver agreement of ERP to diagnose AIP were 71, 83 and 0.45%, respectively across all participating centres (table 2). We also computed the

Phase II results

Sensitivity for AIP (95% CI) Specificity for AIP (95% CI) Interobserver agreement (k)

Overall

USA (N[8)

Europe (N[8)

Asia (N[5)

Asia vs rest

71% (66% to 75%) 83% (78% to 86%) 0.45

65% (57% to 72%) 84% (76% to 89%) 0.42

63% (56% to 71%) 83% (76% to 88%) 0.45

91% (83% to 95%) 80% (70% to 87%) 0.52

p<0.05 NS

AIP, autoimmune pancreatitis; NS, non-significant.

668

Gut 2011;60:666e670. doi:10.1136/gut.2010.207951

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas Table 3 Sensitivity and specificity of the four pancreatogram features 1. Long (>1/3 the length of the pancreatic duct) narrow stricture 2. Lack of upstream dilatation from the stricture (<5 mm) 3. Multiple strictures 4. Side branches arising from the stricture site

Sensitivity (%)

Specificity (%)

38

97

62

89

26 66

98 73

individual performance characteristics of the four ERP features identiﬁed in phase I (table 3). Of these four features the presence of a long narrow stricture or multiple strictures was the most speciﬁc ($97%) but the least sensitive (#38%) We then computed the performance characteristics of these four ERP features in various permutations and combinations. The top ﬁve combinations are presented in (table 4).

DISCUSSION AIP is a relatively new and uncommon disease. Kamisawa et al ﬁrst recognised it as the pancreatic manifestation of a systemic ﬁbroinﬂammatory disease called IgG4-associated systemic disease.12e14 It was not until 2002 that the ﬁrst established diagnostic criteria were published.15 Since then many other diagnostic criteria have been published, including the JPS 2006, Mayo Clinic HISORt criteria and the Korean criteria.7 15e17 These criteria rely on a combination of imaging of the pancreas (ERP, CT and MRI), histology of the pancreas, serology, other organ involvement (bile ducts, salivary glands, retroperitoneum) and a dramatic response to steroid treatment to diagnose AIP. The differences in the speciﬁc criteria often reﬂect regional variations in practice and familiarity with the various diagnostic modalities (eg, ERP vs magnetic resonance cholangiopancreatography (MRCP) in the Japanese vs Korean criteria). The role of an ERP Figure 4 Differentiating autoimmune pancreatitis (AIP) from pancreatic cancer (PaC). H1, histology 1 (pancreatic core biopsy or resection specimen) highly suggestive/diagnostic of AIP; H2, histology 2 (pancreatic core biopsy or resection specimen) supportive of AIP; I-CT1, imaging criterion 1 (imaging of pancreas CT/MRI) highly suggestive/diagnostic of AIP; I-CT2, imaging criterion 2 (imaging of pancreas CT/MRI) supportive of AIP; I-P1, imaging of the pancreas 1 (pancreatogram) highly suggestive/ diagnostic of AIP; I-P2, imaging of the pancreas 2 (pancreatogram) supportive of AIP; OOI-1, other organ involvement highly suggestive/diagnostic of AIP; OOI-2, other organ involvement supportive of AIP; Rt, response to steroid treatment; S-1, serology highly suggestive/diagnostic of AIP; S-2, serology supportive of AIP.

Table 4 Sensitivity and specificity to diagnose autoimmune pancreatitis using combinations of the four endoscopic retrograde pancreatography features Criteria*

Sensitivity (%)

Specificity (%)

1 and 2 1 or 2 2 or 3 1, 2, 3 and 4 1, 2, 3,or 4

47 78 89 52 100

100 91 91 91 66

*The four features are: (1) long (>1/3 the length of the pancreatic duct) narrow stricture; (2) lack of upstream dilatation from the stricture (<5 mm); (3) multiple strictures; and (4) side branches arising from the stricture site.

in the diagnosis of AIP is often debated.18 The Japanese criteria mandate the use of an ERP to diagnose AIP, the Korean criteria allow use of MRCP for ductal imaging, and HISORt criteria do not mandate pancreatography (MCRP or ERP).19 To date, the performance characteristics of ERP in AIP have not been systematically studied. In phase I of the study we found that the performance characteristics of ERP to diagnose AIP were poor, with signiﬁcant differences between Asian and non-Asian physicians. While Asian physicians performed very well in terms of sensitivity and speciﬁcity, the sensitivity of most non-Asian physicians was poor. Even though endoscopic retrograde cholangiograms (ERCs) are frequently performed in the West in patients with obstructive jaundice, injection of the pancreatic duct (ERP) is carefully avoided to minimise the risk of pancreatitis. Thus, we believe the inability of non-Asian readers to identify AIP on ERP is mainly due to lack of familiarity with ERP changes in AIP. This represents an example of ‘the eye does not see what the mind does not know’. There were four physicians (not all from Asia) who were highly adept at diagnosing AIP solely based on the ERP features. They identiﬁed four features (ﬁgures 2 and 3) on ERP which

Patients presenting with obstructive jaundice and/or pancreatic mass Pancreatic findings on CT/MRI: Stratify patients into Imaging Groups 1, 2 or 3 Imaging Highly Suggestive of AIP (I-CT1) Serology and Other Organ Involvement

Imaging Indeterminate (I-CT2) OOI present

Cancer work-up negative

One supportive or highly suggestive feature of AIP (S1,S2 or O1,O2)

Yes

Cancer work-up

No

Imaging Highly Suggestive of PaC Consider managing as PaC unless OOI suggestive of AIP (O2) on imaging

Pancreatogram, Serology, Other Organ Involvement One highly suggestive or two supportive features of AIP (I-P1or 2, S1or 2 or O1or 2)

•None or only one supportive feature of AIP (I-P2 or S2 or O2) Pancreatic core biopsy

Highly suggestive/Supportive of AIP Treat with prednisone 40 mg/day X 2 weeks Reassess serum IgG4, CA 19-9, pancreatic morphology Response to steroids (RT) No Reconsider diagnosis

Positive for cancer

Inconclusive/Not performed Steroid Trial Not performed Surgical resection

Yes AIP diagnosis confirmed

No OOI

Diagnostic of AIP

Positive for cancer Manage as PaC

Modified from Chari ST, Takahashi N, Levy MJ, et al A Diagnostic Strategy to Distinguish Autoimmune Pancreatitis from Pancreatic Cancer. Clinical Gastroenterology and Hepatology 2009;7(10):1097-1103.

Gut 2011;60:666e670. doi:10.1136/gut.2010.207951

669

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas they found helpful to diagnose AIP. We asked the next logical question: Can these features help other physicians to identify AIP on an ERP? Thus a teaching module illustrating the four features was created. In phase II of the study all participating centres and physicians showed an improvement in sensitivity with a modest, but statistically insigniﬁcant, drop in speciﬁcity. None of the four ERP features by themselves was diagnostic of AIP. The presence of all four features was highly speciﬁc (91%) but only modestly sensitive (52%) for AIP (table 4). The presence of pancreatic duct strictures (single or multiple) without upstream dilatation (<5 mm) had the highest speciﬁcity for AIP (>90%). These data suggest that there are features on an ERP which are speciﬁc for AIP, and we believe that making readers aware of these features can augment their ability to diagnose AIP. We have observed a similar phenomenon in regards to histological features of AIP. Until the histological features of AIP were highlighted by observant pathologists, it was generally assumed that the aetiology of CP could not be ascertained by histology.2 3 10 The main strengths of our study are its designdthat is, it was a randomised, blinded, multicentre study with many physicians participating. This study demonstrates that knowledge from a few experts is potentially transferable to others. The limitation of our study is that readers were not provided with any other information about the patients, including ﬁndings on the cholangiogram, but this was by choice. Supplementary tables 3 and 4 illustrate the clinical proﬁle of the patients with AIP who participated in our study. In real life, physicians would potentially have access to such patient information, including clinical presentation and ﬁndings on CT scan and cholangiograms. However, despite the lack of such data, adept readers were able to identify AIP with a high degree of sensitivity. Our study shows that there are speciﬁc ERP features of AIP which are potentially useful. However, since physicians in the West do not perform pancreatograms in the setting of obstructive jaundice, where does ERP ﬁt in the strategy to distinguish AIP from pancreatic cancer? Recently two diagnostic strategies for distinguishing AIP from PaC have been published and compared.2 3 10 One relies on ERP and CT features2 and the other uses CT features but does not use ERP at all.3 We believe the ideal strategy would be to use a combination of these strategies, with the use of ERP being tailored to ﬁndings on cross-sectional imaging. If there is diffuse ‘sausage-shaped’ enlargement of the pancreas without ductal dilatation in the presence of elevated serological markers of AIP, especially IgG4, then pancreatography probably has little added beneﬁt for the diagnosis of AIP.2 However, if ﬁndings on cross-sectional imaging are not typical, ERP may be helpful in identifying AIP in those in whom a diagnosis of PaC is suspected and the other studies are non-diagnostic. Based on these principles, we propose an algorithm incorporating ERP features to differentiate AIP from pancreatic cancer (ﬁgure 4 and Supplementary ﬁgure 1).

670

In summary, this multicentre study shows that the ability to diagnose AIP based on ERP features alone is limited but can be improved with the knowledge of some key radiographic features. We have identiﬁed four key ERP features and determined their performance characteristics in diagnosing AIP. Competing interests None. Ethics approval The study was approved by the Mayo Clinic’s institutional review board. Contributors Study concept and design, acquisition of data, analysis and interpretation of data, drafting of the manuscript, critical revision of the manuscript for important intellectual content, statistical analysis and study supervision: AS, GW and STC. Study participant, critical revision of the manuscript for important intellectual content, drafting of the manuscript: MJL, TK, GW, THB, NT and MDT. Study participant and drafting of the manuscript: NIC, JEC, N, GJJ, KO, RKP, SPP, BTP, SR, RPS, NSS, KU and SSV. Analysis and interpretation of data, drafting of the manuscript: FE. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19.

Sugumar A, Chari S. Autoimmune pancreatitis: an update. Expert Rev Gastroenterol Hepatol 2009;3:197e204. Chari ST, Takahashi N, Levy MJ, et al. A diagnostic strategy to distinguish autoimmune pancreatitis from pancreatic cancer. Clinl Gastroenterol Hepatol 2009;7:1097e103. Kamisawa T, Imai M, Yui Chen P, et al. Strategy for differentiating autoimmune pancreatitis from pancreatic cancer. Pancreas 2008;37:e62e7. Yoshida K, Toki F, Takeuchi T, et al. Chronic pancreatitis caused by an autoimmune abnormality. Proposal of the concept of autoimmune pancreatitis. Dig Dis Sci 1995;40:1561e8. Okazaki K, Kawa S, Kamisawa T, et al. Clinical diagnostic criteria of autoimmune pancreatitis: revised proposal. J Gastroenterol 2006;41:626e31. Chari ST. Diagnosis of autoimmune pancreatitis using its five cardinal features: introducing the Mayo Clinic’s HISORt criteria. J Gastroenterol 2007;42(Suppl 18):39e41. Chari ST, Smyrk TC, Levy MJ, et al. Diagnosis of autoimmune pancreatitis: the Mayo Clinic experience. Clin Gastroenterol Hepatol 2006;4:1010e16; quiz 934. Axon AT, Classen M, Cotton PB, et al. Pancreatography in chronic pancreatitis: international definitions. Gut 1984;25:1107e12. Sarner M, Cotton PB. Classification of pancreatitis. Gut 1984;25:756e9. Sugumar A, Chari ST. Distinguishing pancreatic cancer from autoimmune pancreatitis: a comparison of two strategies. Clin Gastroenterol Hepatol 2009;7(11 Suppl):S59e62. Kruse A, Thommesen P, Frederiksen P. Endoscopic retrograde cholangiopancreatography in pancreatic cancer and chronic pancreatitis. Differences in morphologic changes in the pancreatic duct and the bile duct. Scand J Gastroenterol 1978;13:513e17. Kamisawa T, Okamoto A. Autoimmune pancreatitis: proposal of IgG4-related sclerosing disease. J Gastroenterol 2006;41:613e25. Kamisawa T. IgG4-related sclerosing disease. Intern Med 2006;45:125e6. Kamisawa T, Funata N, Hayashi Y, et al. A new clinicopathological entity of IgG4-related autoimmune disease. J Gastroenterol 2003;38:982e4. Diagnostic criteria for autoimmune pancreatitis by the Japan Pancreas Society. J Jpn Pancreas (Suizou) 2002;17:587. Kim KP, Kim MH, Kim JC, et al. Diagnostic criteria for autoimmune chronic pancreatitis revisited. World J Gastroenterol 2006;12:2487e96. Otsuki M, Chung JB, Okazaki K, et al. Asian diagnostic criteria for autoimmune pancreatitis: consensus of the JapaneKorea symposium on autoimmune pancreatitis. J Gastroenterol 2008;43:403e8. Kwon SMD, Kim M-H, Choi EK. The diagnostic criteria for autoimmune chronic pancreatitis: it is time to make a consensus. Pancreas 2007;34:279e86. Forsmark C. Chronic pancreatitis. In: Feldman M, Friedman LS, Brandt LJ, eds. Sleisenger & Fordtran’s Gastrointestinal and Liver Disease. 8th edn. Philadelphia, USA: Saunders, Elsevier, 2008:1272e3.

Gut 2011;60:666e670. doi:10.1136/gut.2010.207951

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas

Panhematin provides a therapeutic benefit in experimental pancreatitis Aida Habtezion,2 Raymond Kwan,1 Ehsaan Akhtar,2 Stephen P Wanaski,3 Stephen D Collins,3 Ronald J Wong,4 David K Stevenson,4 Eugene C Butcher,5 M Bishr Omary1 See Commentary, p 569 < Additional figures and table

are published online only. To view these files please visit the journal online (http://gut.bmj. com). 1

University of Michigan Medical School, Department of Molecular & Integrative Physiology, Ann Arbor, Michigan, USA 2 Stanford University School of Medicine, Department of Medicine, Division of Gastroenterology and Hepatology, Stanford, California, USA 3 NeuroTherapeutics Pharma, Inc., Chicago, Illinois, USA 4 Stanford University School of Medicine, Department of Pediatrics, Stanford, California, USA 5 Stanford University School of Medicine, Department of Pathology, Stanford, California, USA Correspondence to Dr Aida Habtezion, Stanford University School of Medicine, Department of Medicine, Division of Gastroenterology & Hepatology, 300 Pasteur Drive, Stanford, CA 94305, USA; [email protected] Revised 31 August 2010 Accepted 29 September 2010 Published Online First 15 December 2010

ABSTRACT Background and aim Acute pancreatitis (AP) can result in pancreatic necrosis and inflammation, with subsequent multi-organ failure. AP is associated with increased neutrophil recruitment and a rise in pro-inflammatory cytokines such as TNFa. Pretreatment with haemin, results in recruitment of haem-oxygenase-1 (HO-1)+ macrophages and protects against experimental pancreatitis. It is not clear whether modulation of HO-1 after onset of disease has a protective role. In this study, we tested the utility of Panhematin, a water-soluble haemin formulation, in activating and inducing pancreatic HO-1, and as a therapeutic agent in treating mouse acute pancreatitis. Methods We defined the distribution of radiolabelled haemin, then used in vivo HO-1eluciferase bioluminescence imaging and the CO-release assay to test Panhematin-induced upregulation of HO-1 transcription and activity, respectively. Using two welldefined AP murine models, we tested the therapeutic benefit of Panhematin, and quantified cytokine release using a luminex assay. Results Intravenously administered Panhematin induces rapid recruitment of HO-1+ cells to the pancreas within 2 h and de novo splenic HO-1 transcription by 12 h. Despite high baseline spleen HO-1 activity, the pancreas is particularly responsive to Panhematin-mediated HO-1 induction. Panhematin-treated mice, at various time points after AP induction had significant reduction in mortality, pancreatic injury, together with upregulation of HO-1 and downregulation of pro-inflammatory cytokines and CXCL1, a potent neutrophil chemoattractant. Conclusions Despite AP-associated mortality and morbidity, no effective treatment other than supportive care exists. We demonstrate that Panhematin leads to: (i) rapid induction and activation of pancreatic HO-1 with recruitment of HO-1+ cells to the pancreas, (ii) amelioration of AP even when given late during the course of disease, and (iii) a decrease in leucocyte infiltration and pro-inflammatory cytokines including CXCL1. The utility of Panhematin at modest doses as a therapeutic in experimental pancreatitis, coupled with its current use and safety in humans, raises the potential of its applicability to human pancreatitis.

INTRODUCTION Acute pancreatitis (AP) accounts for over 220 000 hospital admission every year in the United States alone, with approximately 20% of these patients suffering a severe disease course and can be associated with signiﬁcant morbidity and mortality.1 2 AP is thought to develop as a result of an injury to Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

Significance of this study What is already known about this subject?

< Animal studies in acute pancreatitis (AP) show that

HO-1 upregulation by haemin, prior to induction of pancreatitis, prevents pancreatic injury.

What are the new findings?

< Panhematin, a water-soluble formulation of

haemin that is used to treat porphyrias, induces rapid upregulation of pancreatic HO-1 within 2 h of dosing. < Panhematin offers therapeutic benefit not only when given early during AP course but also when give late during the course of experimental pancreatitis. < The therapeutic benefit of Panhematin is associated with reduction in pro-inflammatory cytokines and in a potent neutrophil chemoattractant (CXCL1). Neutrophils play a central role in disease pathogenesis and depletion of neutrophils is already known to reduce AP in experimental models.

How might it impact on clinical practice in the foreseeable future? < Despite significant morbidity and mortality associated with AP, no effective treatments exist aside from supportive care such as fluid resuscitation. The therapeutic benefit of Panhematin in experimental pancreatitis raises the hypothesis that pharmacological administration of Panhematin may ameliorate disease severity in patients with AP.

the pancreatic acini, leakage and activation of pancreatic enzymes within the tissue, and the subsequent initiation of autodigestion and pancreatic injury.3e5 The most feared complication of AP and the local inﬂammatory response are systemic and multiple organ failure, which are associated with high mortality.2 6 Despite the burden of disease, aside from supportive management, no effective therapy exists for treating AP.7 Thus, a search for new treatment modalities for AP remains an important goal in the ﬁeld. Haem-oxygenase-1 (HO-1) is a stress inducible enzyme, while its related isoform HO-2 is constitutively expressed.8 HOs are crucial rate-limiting enzymes that cleave haem into iron, carbon monoxide (CO) and biliverdin.9 10 We previously 671

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas showed that pre-treatment with haemin, a haemoglobin prosthetic moiety that upregulates HO-1, protects mice from acute experimentally induced-pancreatitis using two independent pancreatic injury models.11 The protective role of haemin is mediated via HO-1+ F4/80+ cells that are recruited to the pancreas.11 Given the utility of haemin as a prophylactic agent for pancreatitis development, we hypothesised that haemin may offer a therapeutic beneﬁt in experimental pancreatitis. Haemin is soluble at alkaline pH, and therefore would have to be given intraperitoneally (i.p.) in mice already experiencing signiﬁcant peritoneal inﬂammation. Given the limitation of haemin administration via the i.p. route, the use of Panhematin (PH), a water soluble intravenous formulation of haemin, is a clinically attractive possibility, although protocols for reconstitution of haematin for intravenous administration are available.12 Notably, PH is an FDA-approved drug for the treatment of acute intermittent porphyria.13 14 In the study herein, we show that PH is an effective therapeutic in two experimental models of acute pancreatitis. There is early upregulation of HO-1 and recruitment of HO-1+ cells to the pancreas following PH administration. Although haemin administration does not result in its preferential accumulation in the pancreas, there is an increase in HO-1 protein expression and signiﬁcant upregulation in HO-1 activity in the pancreas after PH therapy. Consistent with the anti-inﬂammatory effects of PH, there is a decrease in pancreatic and serum tumour necrosis factor a (TNFa), interleukin 12p40 (IL-12p40) and IL-6. Neutrophils and their recruitment to the inﬂamed pancreas have been shown to play a central role in the pathogenesis of AP.15 We show here that PH treatment leads to a decrease in pancreatic leucocyte inﬁltration and CXCL1, a potent neutrophil chemoattractant, thereby providing a potential therapeutic mechanism for PH against pancreatitis.

MATERIALS AND METHODS Mice treatment and induction of experimental pancreatitis FVB/N mice were purchased from Taconic (Hudson, New York, USA). Young female mice (15e19 g) were fasted for 12 h and then fed a choline-deﬁcient diet (CDD; Harlan Teklad, Madison, Wisconsin, USA) supplemented with 0.5% DL-ethionine (SigmaAldrich, St Louis, Missouri, USA) or normal chow (control group) for 3 days. In addition, an L-arginine model of pancreatitis was carried out using two intraperitoneal doses of L-arginine (given at time 0 and 1 h) in C56BL/6 mice (Taconic) as described previously.16 Mice received PH (Ovation Pharmaceuticals, Deerﬁeld, Illinois, USA) dissolved in sterile water (reported as equivalent to haemin dose, as described on the drug nomogram, per mouse weight) intravenously (i.v.) at various times (supplementary table 1). PH was given i.v. 24 h following the second dose of Larginine. Six- to 8-week-old HO-1-luc transgenic mice, whose transgene contains the full-length (15 kb) HO-1 promoter fused to the reporter gene luciferase (luc) were used for in vivo bioluminescence imaging of HO-1 transcriptional activity.17

Western blotting Mouse pancreata were isolated and frozen immediately in liquid nitrogen. Frozen pancreata were homogenised using a Dounce and 3% SDS/5 mM EDTA/0.187 M Tris-HCl (pH 6.8). Protein concentrations were determined by the bicinchoninic acid (BCA) protein assay. Equal amounts of protein were separated on SDSPAGE followed by electrophoretic transfer onto polyvinylidene ﬂuoride (PVDF) membranes. Anti-HO-1 and anti-HO-2 polyclonal antibodies were purchased from Assay Designs (Ann Arbor, MI, USA). 672

Immunofluorescence and histology Mouse pancreata were frozen in Tissue-Tek OCT compound and sectioned using a cryostat. Slides were ﬁxed in acetone and incubated with antibodies to HO-1 to visualise HO-1+ cells. To assess pancreatic injury, mouse pancreata were ﬁxed in 10% neutral buffered formalin (pH 6.8e7.2). Pancreata were embedded in parafﬁn, sectioned and stained with H&E. The histological score and grading were based on oedema, inﬂammation, haemorrhage and parenchymal necrosis as described previously.11 To assess leucocyte inﬁltration, parafﬁn-embedded sections were stained with biotin-labelled anti-mouse CD45 antibody (BD Biosciences, San Diego, California, USA) followed by streptavidinehorseradish peroxidase (HRP) conjugate and 3,39 -diaminobenzidine (DAB) chromogen solution (R&D Systems, Minneapolis, Minnesota, USA).

Biochemical analysis Blood was collected from mice by cardiac puncture, and serum was isolated from these samples for subsequent lipase and blood urea nitrogen (BUN) level determination using Beckton Dickinson Microtainer Serum Separator Tubes (Beckton Dickinson, Franklin Lakes, New Jersey, USA). Pancreatic trypsin activity was determined as described previously.11

Bacterial culture Blood, pancreatic and liver tissues were collected using sterile technique and submitted for bacterial growth analysis to the our institutional microbiology lab. Pancreas and liver were ﬁrst homogenised in sterile saline. 55

Fe-haemin and tissue homogenisation

The indicated amounts of 55Fe-haemin (RI Consultants, Chicago, Illinois, USA). were injected intravenously into mice. Mice were sacriﬁced at various times, and brain, heart, kidney, pancreas, small intestines, stomach, colon, liver, and spleen were isolated, weighed, and homogenised as described for immunoblotting. Scintillation counts of homogenised tissues were obtained using a Beckman Coulter counter (Beckman Coulter, Brea, California, USA).

Tissue sectioning and autoradiography Whole mice were embedded in Tissue-Tek OCT compound and sectioned using a Leica CM3600 Macrotome. Sections were placed on Kodak Biomax XAR ﬁlm and exposed for 1 month in order to visualise 55Fe-haemin tissue distribution.

HO-1 promoter activity using in vivo bioluminescence imaging HO-1-luc transgenic mice with an FVB/N background received one dose of PH on day 0.11 17 Whole body imaging of luc expression (hence HO-1 promoter activity) of the PH- or VE-treated HO-1-luc mice were performed at 0, 2, 4, 6, 8, 12, 18, 24 and 30 h post-PH treatment. Luciferin (100 ml, 30 mg/ml) per 20 g of mouse weight (ﬁnal 150 mg/kg body weight) was administered i.p. to each mouse 10 min prior to imaging. Mice were then placed in an In Vivo Imaging System (IVIS; Alameda, California, USA) and the photons emitted from the tissues were quantiﬁed using LivingImage software v3.2 (Caliper Life Sciences, Alameda, CA, USA).

Haem-oxygenase activity Tissues were harvested and prepared as follows: 100620 mg of tissue was placed in a 1.5 ml microfuge tube and nine volumes of phosphate buffer was added, diced, and immediately sonicated at 08C using an ultrasonic cell disruptor with an ice-cold 3 mm microprobe yielding a concentration of 2 mg fresh tissue weight Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas per 20 ml of sonicate. Tissue HO activity was determined through measurement of CO as previously described.18 Brieﬂy, 20 ml of tissue sonicate was incubated at 378C for 15 min in COfree septum-sealed 2 ml phials, containing 20 ml of methemalbumin (150 mmol haem/l:15 mmol BSA/L) and 20 ml of NADPH (4.5 mmol/l). Blank reaction phials contained 20 ml of phosphate buffer in place of NADPH. The reaction was stopped by the addition of 5 ml of 15% sulfosalicylic acid.18 The CO produced and released into the phial headspace was then quantiﬁed by gas chromatography. HO activity was calculated as pmol CO generated/h/mg wet weight.

Luminex assay The Luminex assay was performed as recommended by the manufacturer (Panomics/Affymetrix, Santa Clara, Califonia, USA). Assays were performed in duplicate using the Luminex 200 IS System (Luminex Corporation, Austin, Texas, USA). Individual cytokines and chemokines were identiﬁed and classiﬁed by the red laser, and levels were quantiﬁed using the green laser. Digital images of the bead array were captured following laser excitation and were processed on a computer workstation. Standard curves and reports of unknown samples were prepared using BeadView and MiraiBio software.

In situ hybridisation In situ hybridisation of pancreatic sections was performed as described.19 In brief, digoxigenin-labelled sense and anti-sense RNA probes were generated by PCR ampliﬁcation with the T7 promoter incorporated into the primers. The primers used for CXCL1 were: 59 -ATGATCCCAGCCACCCGCTCGCTT-39 (sense) and 59 -CCGTTACTTGGGGACACCTTTTAGCATC-39 (antisense). In vitro transcription was performed with a digoxigenin RNA-labelling kit and T7 polymerase (Roche Diagnostics, Indianapolis, Indiana, USA). Pancreatic sections from parafﬁn blocks were used for hybridising with either sense or anti-sense probes as described.19 Sections were developed with diaminobenzidine (GenPoint Kit; DAKO, Carpinteria, California, USA) and counterstained in haematoxylin.

Statistical analysis The Student t test was used to assess statistical signiﬁcance and a p value of less than 0.05 was considered signiﬁcant. Values are expressed as mean6SEM. Unless indicated, results are from at least three independent experiments. Survival curve comparison was based on the log-rank (ManteleCox) test.

RESULTS Systemic treatment with PH results in early upregulation of HO-1 and recruitment of HO-1+ cells to the pancreas We previously showed that multiple i.p. injections of haemin during a 1-week period upregulate HO-1 signiﬁcantly by days 4 and 5 of starting the injections.11 Here, we examined whether HO-1 expression can be upregulated with a single i.p. or intravenous dose of PH to ascertain whether PH can be utilised for AP therapy. Interestingly, we found that HO-1 expression was upregulated as early as 2 h following a single-dose administration of PH (ﬁgure 1A). Similarly, immunoﬂuorescence staining revealed increasing numbers of HO-1+ cells at 4 h post-injection, thereby suggesting rapid recruitment of HO-1+ cells into the pancreas regardless of the route of PH administration (ﬁgure 1B). 55

Fe-haemin concentrates preferentially in the liver and spleen

Previous studies using intravenous 55Fe-haematin in rhesus monkeys showed a half-life in plasma ranging from 2.2 to Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

Figure 1 Induction of haem-oxygenase-1 (HO-1) by Panhematin (PH). (A) Western blots were performed on pancreatic lysates from mice at time 0, 2 or 4 h (h), after receiving intraperitoneal (IP) or intravenous (IV) PH. Note, PH upregulates HO-1 (upper band, arrow) in the pancreas as early as 2 h whereas HO-2 remains unchanged. (B) Frozen pancreas sections at 2 h and 4 h following treatment with PH were stained with anti-HO-1 (red) antibody. Note, PH causes influx of HO-1+ cells into the pancreas. 4.8 h.20 The majority of the injected radioactivity was observed in the liver (48.2%) 21 h after 55Fe-haematin injection, but haematin distribution to the pancreas was not examined.20 Thus, based on the beneﬁcial effects of haemin in preventing pancreatitis and the early upregulation of HO-1 protein expression (ﬁgure 1A) following a single dose of PH, we investigated the distribution of 55Fe-haemin in homogenised pancreata and other tissues 4 and 24 h after intravenous injection.11 Consistent with previous studies, liver, spleen and colon (in decreasing amounts) had the highest accumulation of radioactivity relative to other tissues at 24 h (ﬁgure 2A). 55Fehaemin accumulation in the pancreas was limited as compared with the three major organs of radiolabelled haemin deposition. Similarly, whole mouse section autoradiography demonstrated high activities in the liver and spleen (ﬁgure 2B).

Panhematin given 8 and 24 h following CDD initiation, or after L-arginine administration, attenuates pancreatic injury Pre-treatment with haemin protects against mild to moderate (caerulein model) and severe (CDD model) forms of experimental pancreatitis.11 In order to evaluate whether PH can treat early or late pancreatitis, we used the CDD-induced pancreatitis model, where mice are fed CDD for 3 days thereby resulting in severe haemorrhagic pancreatitis with peak injury on day 3. Lipase levels in CDD-induced pancreatitis correlate well with pancreatic injury as assessed by histological scores.11 We determined the expression of HO-1 and extent of pancreatic injury during CDD feeding over time. After 8 and 24 h of CDD feeding alone, there was no notable HO-1 up-regulation (ﬁgure 3A; lanes 1, 2, 7, 8). A signiﬁcant increase in serum lipase was observed at 24 h, but not after 8 h (ﬁgure 3B). 673

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas Figure 2 55Fe-haemin tissue distribution. (A) FVB/N mice were intravenously injected with 2 mCi of 55 Fe-labelled haemin. The haemin dosage was adjusted to 6 mg/g with cold PH. After 4 and 24 h, mice were euthanised, and the specified organs were isolated. A portion of each organ was homogenised and subjected to scintillation counting. The data is presented as cpm/mg of wet weight tissue. Each value represents an average of three mice. Note that 55Fehaemin does not concentrate in the pancreas. (B) Similarly mice that received 2 mCi of 55Fe-labelled haemin as in (A) were euthanised after 24 h and quickly frozen and embedded in TissueTek OCT compound. Whole body sections (50 mm) were placed on a piece of tape and developed on Kodak Biomax XAR film. Two individual mouse images (a and c) are shown with their corresponding autoradiograph (b and d, respectively). Liver and spleen (arrows) show high activity.

We then assessed pancreatic HO-1 expression after 72 h of CDD feeding in mice treated with either vehicle (VE) or PH at 8 or 24 h after initiating CDD. Notably, a single dose of PH given either at 8 or 24 h following initiation of CDD induced and maintained HO-1 upregulation at 72 h (ﬁgure 3A; lanes 5, 6, 11, 12). This led us to test whether this dosing regimen after feeding with CDD offered a therapeutic beneﬁt. We found that a single dose of PH, equivalent to a haemin dose of 25 mg/g mouse weight, which previously conferred protection against AP,11 dramatically suppressed serum lipase elevation (ﬁgure 3C) and pancreatic injury as determined after scoring of the damage using deﬁned histological criteria (ﬁgure 3D,E) and trypsin activation (ﬁgure 3F) at 72 h. In addition, mice receiving PH at 8 and 24 h had better survival at 72 h (ﬁgure 3G). Importantly, PH was also similarly effective in treating L-arginine induced pancreatitis (ﬁgure 4). Blood, pancreas and liver homogenate cultures did not yield any bacterial growth in either 24 h VE or PH treated mice euthanised at 72 h following initiation of CDD (data not shown).

Late Panhematin treatment during the course of acute pancreatitis requires dose adjustment Given the therapeutic effects of administering PH 8 and 24 h after CDD feeding, we investigated whether PH can reverse or treat late AP. As expected, there was further rise in serum lipase at 36 and 48 h following CDD initiation (ﬁgure 5A). However, PH at the dose of haemin (equivalent to 25 mg/g weight) given at 36 and 48 h did not protect against AP (ﬁgure 5B). The difference 674

in the protective effect of PH at 36 and 48 h versus 8 and 24 h prompted us to examine whether the dose of PH administered caused potential toxic effects similar to that of high dose haemin,21 particularly in severely ill mice during late phase pancreatitis. Consistent with this notion, PH treated mice have higher serum BUN levels (supplementary ﬁgure 1). Thus, we examined upregulation of pancreatic HO-1 at lower doses of PH to determine if we could mitigate potential PH toxicity. Dose response studies showed that administration of PH at 6.2 mg/g of mouse weight afforded signiﬁcant HO-1 overexpression in the pancreas (ﬁgure 6A). Based on this ﬁnding and the recommended upper limit dosage in humans (equivalent to 6 mg/kg body weight over 24 h14;), we used 6 mg/g weight dosing to test PH effectiveness in late AP. Notably, this lower dose of PH was effective in treating AP when given at 48 h, but not earlier at the 24 h time point (ﬁgure 6B). For comparison, serum lipases from controls (no CDD) and mice fed CDD only for the ﬁrst 24 or 48 h are shown (ﬁgure 6C).

Effect of Panhematin on haem-oxygenase-1 activity as assessed by in vivo bioluminescence imaging and CO release In order to address the mechanism of PH action, we studied whether there is de novo transcription of HO-1 and whether the induction of HO-1 by PH correlates with HO activity by measuring HO-1 promoter activity using HO-1-luc transgenic mice (Luc-HO-1+/+) and via release of CO from homogenised tissues, respectively. Mice were imaged 0, 2, 4, 6, 8, 12, 18, 24 and 30 h following a single dose of either VE or PH. Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas Figure 3 Effect of Panhematin (PH) in experimental pancreatitis. (A) Female mice on a FVB/N background were fed a choline deficient diet (CDD) to induce acute pancreatitis. Some mice were fed CDD but did not receive PH or vehicle (VE) and were euthanised at 8 or 24 h. Other mice were fed CDD for 72 h and treated with a single intravenous dose of PH or VE at 8 or 24 h following the initiation of CDD. Total pancreas lysates were analysed by SDS-PAGE and western blotting for haem-oxygenase-1 (HO-1) and HO-2 proteins. (B) Mice were fed either normal diet (white bar) or CDD for 8 (grey bar) and 24 h (black bar) and then euthanised. Serum was collected for lipase determination. (C) Some mice received either vehicle (VE; white bar) or PH (black bar) treatment at 8 h following initiation of CDD and other group of mice received either VE (white bar) or PH (black bar) treatment at 24 h following initiation of CDD and were then euthanised at 72 h. Serum was used for determination of lipase levels at 72 h. (D) Mice were fed CDD for 72 h and euthanised. The mice were treated with either vehicle (VE) control (a) or PH (b) at 24 h following the initiation of CDD. Pancreata were isolated and fixed in 10% formalin, and sections were stained with H&E. Black and green arrows indicate areas of haemorrhage and necrosis respectively. (E) Histological scores of mice treated with VE or PH at either 8 h or 24 h after the initiation of CDD are shown. Mice were euthanised at 72 h and pancreas processed as in D and scored. (F) Pancreatic tissues from VE or PH treated mice at 24 h after the initiation of CDD and euthanised at 72 h were used to determine trypsin (pmol/mg protein) enzyme activity. *p<0.05. Results shown are from four independent experiments with a total of 6e8 mice per group. (G) Survival curve from 8 and 24 h (combined for the two time points) VE or PH treated mice following the initiation of CDD is shown. Results shown are from four independent experiments with total of 22 mice per group. *p<0.05 using log-rank (ManteleCox) test.

Signiﬁcant bioluminescence was observed at 12 h on the left dorsal and lateral areas in mice treated with PH (ﬁgure 7A), likely corresponding to the spleen. Based on this ﬁnding, we then treated normal FVB/N mice (similar to those used in induction of AP) with either VE or PH, and assayed HO activity in lung, liver, spleen, and pancreas sonicates after 12 and 24 h. There was signiﬁcant increase in HO activity in the liver, spleen, and pancreas, but not lung of PH treated mice at both 12 (not shown) and 24 h (ﬁgure 7B). Interestingly, a higher fold increase in HO activity was observed in the pancreas as compared to the spleen following PH treatment (grey bars; ﬁgure 7B).

Panhematin suppresses pancreatic pro-inflammatory cytokines and a potent neutrophil chemoattractant Activated leucocytes during the early phase of pancreatitis release cytokines, which in turn mediate and enhance the inﬂammatory cascade observed in AP. Particularly, TNFa and IL-1 are among the most prominent and well-studied in AP.6 22 23 We performed Luminex assays to determine if PH treatment affects cytokine levels in the pancreas. Although no difference in IL-1 pancreatic levels was noted between 24 h VE or PH treated mice fed with Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

CDD for 60 h, there was a signiﬁcant decrease in TNFa in PH treated mice (ﬁgure 8A). Nevertheless, we could detect a signiﬁcant decrease in IL-1b in serum of PH-treated mice (supplementary ﬁgure 2). In addition, there were signiﬁcant decreases in pancreatic IL-12p40 and IL-6 in the PH-treated mice. We previously showed that single dose of haemin given i.p., in non-CDD fed mice, increased pancreatic CCL2 (MCP-1) and CCL3 (MIP-1a) but not CCL5 (RANTES) mRNA 24 h later.11 Herein, we did not detect signiﬁcant differences in pancreatic protein levels of CCL3 or CCL5 after 60 h CDD feeding in mice treated with either VE or PH at 24 h during the feeding (ﬁgure 8B). We elected to harvest the pancreas at 60 h and not 72 h, since all the mice would be alive at 60 h. Based on the fact that CXCL1 is a potent neutrophil attractant and neutrophils play an important role in pancreatic injury during AP,24 we tested whether PH can modulate pancreatic CXCL1. Consistent with the ﬁndings that PH can treat experimental pancreatitis, CXCL1 was signiﬁcantly decreased in mice treated with PH (ﬁgure 8B). In order to determine the cellular source for the increased CXCL1, we performed in situ hybridisation with CXCL1 sense 675

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas

Figure 4 Effect of Panhematin (PH) in L-arginine model of pancreatitis. (A) Mice were either treated with vehicle (VE) or PH at 24 h following the last L-arginine dose (time 1 h) and euthanised 72 h after the first dose of L-arginine (time 0) and serum was collected for lipase determination. (B) Bar graph denotes pancreatic histological scores from the same mice in (A). (C) Representative pancreata isolated from the same mice in (A), fixed in 10% formalin and H&E stained sections are shown. *p<0.05. (ﬁgure 8C, d) and anti-sense (ﬁgure 8C, aec) probes. Relative to VE-treated mice, pancreata from PH-treated mice had signiﬁcantly reduced CXCL1 mRNA levels (ﬁgure 8C, b vs c). In VEtreated mice, the CXCL1 mRNA was found primarily proximal to vascular areas (ﬁgure 8C, b). Consistent with the reduction in pancreatic CXCL1, pancreata from PH-treated mice had decreased leucocyte inﬁltration as demonstrated by a decrease in CD45+ cells (ﬁgure 8D; c and d vs a and b).

DISCUSSION AP remains a challenging clinical problem, particularly in patients with severe disease. Despite a handful of clinical trials with pharmacological agents, no effective treatment exists for AP. We previously demonstrated that pre-treatment, as a prophylactic, with haemin or haemin-activated cells protects against experimentally induced pancreatic injury.11 In the study herein, the beneﬁcial effect of haemin prophylaxis has been extended to therapeutic intervention using the haemin formulation PH in mice undergoing two different models of pancreatitis. Importantly, even at a later time during AP, PH was effective in minimising the observed pancreatic injury. This ﬁnding is likely related to the observation that HO-1 in the pancreas is quickly upregulated and maintained following a single PH dose. Clinically, such potency may be critical since

patients with AP are a heterogeneous group, presenting to the hospital at different times during the course of the disease. The limited accumulation of 55Fe-haemin in the pancreas suggests that the pancreas is not a major site for haemin uptake. This is consistent with our previous ﬁnding that HO-1+ F4/80+ cells are recruited to the pancreas from elsewhere and that these recruited cells provide the source for pancreatic HO-1 and are responsible for the observed protection. Potentially, Panhematin may also affect other organs such as the liver and spleen monocyte/macrophages and suppress the systemic inﬂammatory responses. Furthermore, PH not only upregulates HO-1 at the protein level but also functionally as we found signiﬁcant increases in HO activity in the pancreas. The beneﬁcial effects of HO-1 induction may be mediated via its byproducts, CO and biliverdin/bilirubin, since both have been shown to have antiinﬂammatory and anti-oxidant properties.25e28 In support of our ﬁndings, the CO-releasing molecule-2 (CORM-2) demonstrated a protective effect in a rat model of pancreatitis.29 Potential therapeutic limitations of such compounds lie in their instability and insolubility in aqueous solutions. In addition, our study demonstrated the efﬁcacy of PH against AP at various time points whereas CORM-2 efﬁcacy was demonstrated only at one time point immediately after AP induction. The relative ease of PH administration, its potency in inducing HO-1 expression in the pancreas, and its efﬁcacy using two independent models of AP at different stages of the disease renders it an attractive potential candidate therapeutic agent for AP. PH treatment resulted in signiﬁcant reduction in TNFa, IL-6, and IL-12p40 levels in pancreatic tissue. Serum levels of TNFa are increased in patients with AP and mouse models of the disease, and increased levels are correlated with disease severity.29e31 Blockade of TNFa in experimental mouse models ameliorates AP, indicating the importance of this cytokine in disease pathogenesis.32 The reduction of TNFa levels in the pancreas may be one of the mechanisms by which PH improves AP. IL-12 in conjunction with IL-18 has been shown to induce necrotising AP in obese ob/ob mice.33 In humans, IL-12p40 is consistently increased during the course of AP whereas IL-12p70 was only increased on day 1, but signiﬁcantly decreased on the second to fourth day of illness.34 Our ﬁndings are supportive of these results and show that, 60 h post-CDD-feeding, IL-12p40 are much higher as compared to IL-12p70. In addition, PH treatment leads to a decrease in IL-12p40 and an increase in IL-12p70, although the latter was not statistically signiﬁcant. Increased serum IL-6 levels are also observed in AP patients and in experimental mouse models.35 Although IL-6 does not appear to play an important role in initiating or mediating the inﬂammatory cascade in AP, it is considered a useful predictor of disease severity.35 In support of the therapeutic beneﬁt of PH in

Figure 5 Late Panhematin (PH) treatment during acute pancreatitis (AP). (A) Mice were fed either normal diet (white bar) or a choline-deficient diet (CDD) for 36 (grey bar) and 48 h (black bar) and then euthanised. Serum was collected for lipase determination. (B) Mice were fed CDD for 72 h and then euthanised. Groups of mice received either VE or PH treatment at 36 h or 48 h following initiation of CDD. Serum was used for determination of lipase levels. *p<0.05. Results shown are from four independent experiments with a total of 6e8 mice per group. 676

Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas Figure 6 Effect of incremental and lower dose of Panhematin (PH) on pancreas haem-oxygenase-1 (HO-1) induction and treatment of acute pancreatitis (AP). (A) Western blots were performed on pancreatic lysates from mice at 24 h after receiving either no or 1.6, 3.2, 6.2 and 25.0 mg/g intravenous PH. Note, significant HO-1 induction with 6.2 mg/g PH, which is comparable to the upper end of the dose used in humans. (B) Mice were fed a choline-deficient diet (CDD) for 72 h and received either vehicle (VE) or 6 mg/g PH treatment at 24 h following initiation of CDD. The 48-h mice received either VE or 6 mg/g PH treatment at 48 h following initiation of CDD. Serum was used for determination of lipase levels. (C) Mice were fed either normal diet (white bar) or CDD for 24 (grey bar) and 48 h (black bar) and then euthanised. Serum was collected for lipase determination. The levels of serum lipase in these studies were much higher as compared to earlier displayed measurements (eg, figure 3) due to the use of a different assay. *p<0.05. Results shown are from four independent experiments with a total of 6e8 mice per group. experimental AP, we ﬁnd that IL-6 is signiﬁcantly lower in PH-treated animals as opposed to their VE-treated controls. IL-1 blockade, via its receptor or via IL-1 converting enzyme (ICE),36e40 attenuates experimental pancreatitis. PH here did not alter pancreatic IL-1 levels 60 h after the initiation of CDD

feeding, suggesting that an IL-1 independent mechanism may account for the therapeutic action of PH. However, our results do not rule out a role for IL-1 down-modulation in PH-mediated protection against AP since the effect of PH on IL-1 levels may occur early during the course of pancreatitis. This latter

Figure 7 Effect of Panhematin (PH) on in vivo haem-oxygenase-1 (HO-1) tissue distribution. (A) A single dose of either PH or vehicle (VE) was injected intervenously into HO-1 luc transgenic (HO-1-Luc) mice. Whole body images of anaesthetised mice in dorsal (upper panels) and lateral (lower panels) positions were taken at various time points (shown t¼0, 6, 12 and 24 h). The signal intensity scale bars are shown on the right. (B) A single dose of PH or VE was injected intravenously into female FVB/N mice. Mice were sacrificed at 24 h, and liver, lung, spleen and pancreata were sonicated in phosphatebuffered saline (PBS). Tissue sonicates were incubated at 378C for 15 min with methemalbumin plus NADPH. The CO produced was quantified by gas chromatography. HO activity was calculated as pmol CO generated/h/mg wet weight. Percent control indicates the percent difference between tissues from VE versus PH treated mice. Groups of three mice were used to obtain mean values. *p<0.05.

Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

677

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas

Figure 8 PH suppresses pro-inflammatory cytokines and neutrophil chemoattractant, CXCL1. Pancreata were isolated from FVB/N mice that were fed CDD for 60 h and treated with vehicle (VE; white bars) or PH (black bars) at the 24 h time point. Pancreatic homogenates were used for determining cytokine levels using preconfigured kits, standards, and Luminex reader for quantitative analysis. (B) Similarly, chemokine levels were determined as in (A). *p<0.05. (C) Pancreata were isolated from FVB/N mice that were (a) untreated or (bed) fed CDD for 72 h and treated with (b, d) VE or (c) PH at the 24 h time point and embedded in paraffin. Pancreatic sections from paraffin blocks were used for in situ hybridisation with digoxigenin-labelled (d) sense or (aec) anti-sense RNA probes as described in Material and Methods. Sections were developed with diaminobenzidine and counterstained in haematoxylin. Note the perivascular CXCL1 staining in VE-treated (b) sections and the reduction in staining in PH-treated (c) sections. (D) Pancreata were isolated from FVB/N mice that were fed a choline-deficient diet (CDD) for 72 h and treated with (a, b) VE or (c, d) PH at the 24 h time point and embedded in paraffin. Pancreatic sections from paraffin blocks were stained with CD45 antibody as described in Material and Methods. Scale bar 50 mm for (a) and (b), magnified views are shown in (b) and (d). A representative of three groups of mice is shown. hypothesis is supported by the fact that at 60 h we could detect signiﬁcant IL-1b reduction in the serum (supplementary ﬁgure 2) and IL-1b levels in the pancreas are much higher early during CDD feeding (eg, at 24 h as opposed to 48 h or 60 h, data not shown). For example, CORM-2 suppresses IL-1b levels 12 h after induction of AP.29 Similarly and in the same study, levels of the anti-inﬂammatory cytokine, IL-10, increased 12 h after induction of AP. We did not observe signiﬁcant changes in pancreatic IL-10 levels 60 h after CDD diet feeding in response to PH treatment. Again, this suggests that PH-mediated protection does not depend on IL-10 solely or, alternatively, that the effect of PH on IL-10 levels occurs at an earlier time point. In addition to the above-mentioned cytokines, our ﬁndings show that the neutrophil chemoattractant, CXCL1, is reduced in PH-treated mice. CXCL1, a potent neutrophil chemoattractant similar to IL-8 in humans, is expressed in ﬁbroblasts, macrophages and endothelial cells particularly in response to inﬂammatory stimuli.41 42 Furthermore, marked reduction in severity of pancreatitis is observed with depletion of neutrophils.24 43 The fact that PH suppresses local pancreatic CXCL1 expression 678

supports the notion that PH ameliorates AP by mitigating neutrophil recruitment into the pancreas through reduction in CXCL1 levels. This conclusion is further supported by the reduced leucocyte (CD45+ cells) inﬁltration that was observed following PH treatment. One important aspect of our ﬁndings is the dosing of PH. For example, higher and lower doses of PH were effective if given earlier and later during the course of AP, respectively. Over time during the CDD feeding, mice become more dehydrated, as shown by the increase in BUN levels (supplementary ﬁgure 1). Treatment with higher doses of PH likely compromises renal function further, since larger doses of PH are known to cause reversible renal failure.21 In patients with porphyria, PH is given daily for 3 to 14 days based on the resolution of clinical symptoms and appears to be well tolerated. However, if our ﬁndings are translated to patients with AP, care will need to be taken with the dosing. Collectively, our ﬁndings show that the use of PH offers a novel potential therapeutic strategy for AP. Acknowledgements We thank Jean Chen, Hui Ye Zheng and Evelyn Resurreccion for their technical assistance. We are also grateful to Timothy C Doyle for guidance with whole animal sectioning, Kelli Montgomery and Rui Li for assistance with in Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Pancreas situ hybridisation, and Yael Rosenberg-Hasson for assistance with the Luminex assay.

19.

Funding This work was supported by NIH grants DK073909, DK47918 and by Ovation Pharmaceuticals (to MBO), and NIH Digestive Disease Center grants DK56339 to Stanford University and DK34933 to the University of Michigan.

20.

Competing interests The supplier of Panhematin (formerly called Ovation Pharmaceuticals) provided support for this study and has a pending patent application for the possible use of Panhematin in human pancreatitis. Ethics approval All animal protocols were approved by the Institutional Animal Care and Use Committees.

21. 22. 23. 24.

Contributors Aida Habtezion and Raymond Kwan contributed equally to this study. Provenance and peer review Not commissioned; externally peer reviewed.

25. 26.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

Gravante G, Garcea G, Ong SL, et al. Prediction of mortality in acute pancreatitis: a systematic review of the published evidence. Pancreatology 2009;9:601e14. Whitcomb DC. Clinical practice. Acute pancreatitis. N Engl J Med 2006;354:2142e50. Gaisano HY, Gorelick FS. New insights into the mechanisms of pancreatitis. Gastroenterology 2009;136:2040e4. Pandol SJ, Saluja AK, Imrie CW, et al. Acute pancreatitis: bench to the bedside. Gastroenterology 2007;132:1127e51. Vonlaufen A, Wilson JS, Apte MV. Molecular mechanisms of pancreatitis: current opinion. J Gastroenterol Hepatol 2008;23:1339e48. Papachristou GI. Prediction of severe acute pancreatitis: current knowledge and novel insights. World J Gastroenterol 2008;14:6273e5. Pezzilli R. Pharmacotherapy for acute pancreatitis. Expert Opin Pharmacother 2009;10:2999e3014. Durante W. Heme oxygenase-1 in growth control and its clinical application to vascular disease. J Cell Physiol 2003;195:373e82. Otterbein LE, Soares MP, Yamashita K, et al. Heme oxygenase-1: unleashing the protective properties of heme. Trends Immunol 2003;24:449e55. Sikorski EM, Hock T, Hill-Kapturczak N, et al. The story so far: Molecular regulation of the heme oxygenase-1 gene in renal injury. Am J Physiol Renal Physiol 2004;286:F425e41. Nakamichi I, Habtezion A, Zhong B, et al. Hemin-activated macrophages home to the pancreas and protect from acute pancreatitis via heme oxygenase-1 induction. J Clin Invest 2005;115:3007e14. Anderson KE, Bonkovsky HL, Bloomer JR, et al. Reconstitution of hematin for intravenous infusion. Ann Intern Med 2006;144:537e8. Anderson KE, Collins S. Open-label study of hemin for acute porphyria: clinical practice implications. Am J Med 2006;119:801.e19e24. Siegert SW, Holt RJ. Physicochemical properties, pharmacokinetics, and pharmacodynamics of intravenous hematin: a literature review. Adv Ther 2008;25:842e57. Gukovskaya AS, Vaquero E, Zaninovic V, et al. Neutrophils and NADPH oxidase mediate intrapancreatic trypsin activation in murine experimental acute pancreatitis. Gastroenterology 2002;122:974e84. Dawra R, Sharif R, Phillips P, et al. Development of a new mouse model of acute pancreatitis induced by administration of L-arginine. Am J Physiol Gastrointest Liver Physiol 2007;292:G1009e18. Zhang W, Feng JQ, Harris SE, et al. Rapid in vivo functional analysis of transgenes in mice using whole body imaging of luciferase expression. Transgenic Res 2001;10:423e34. Vreman HJ, Wong RJ, Kadotani T, et al. Determination of carbon monoxide (CO) in rodent tissue: effect of heme administration and environmental CO exposure. Anal Biochem 2005;341:280e9.

Gut 2011;60:671e679. doi:10.1136/gut.2010.217208

27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

West RB, Corless CL, Chen X, et al. The novel marker, DOG1, is expressed ubiquitously in gastrointestinal stromal tumors irrespective of KIT or PDGFRA mutation status. Am J Pathol 2004;165:107e13. Sears DA, Huser HJ. Plasma hematin-binding and clearance in the rhesus monkey. Proc Soc Exp Biol Med 1966;121:111e16. Dhar GJ, Bossenmaier I, Cardinal R, et al. Transitory renal failure following rapid administration of a relatively large amount of hematin in a patient with acute intermittent porphyria in clinical remission. Acta Med Scand 1978;203:437e43. He S, Wang L, Miao L, et al. Receptor interacting protein kinase-3 determines cellular necrotic response to TNF-alpha. Cell 2009;137:1100e11. Granger J, Remick D. Acute pancreatitis: models, markers, and mediators. Shock 2005;24(Suppl 1):45e51. Frossard JL, Saluja A, Bhagat L, et al. The role of intercellular adhesion molecule 1 and neutrophils in acute pancreatitis and pancreatitis-associated lung injury. Gastroenterology 1999;116:694e701. Bilban M, Bach FH, Otterbein SL, et al. Carbon monoxide orchestrates a protective response through PPARgamma. Immunity 2006;24:601e10. Morse D, Pischke SE, Zhou Z, et al. Suppression of inflammatory cytokine production by carbon monoxide involves the JNK pathway and AP-1. J Biol Chem 2003;278:36993e8. Otterbein LE, Bach FH, Alam J, et al. Carbon monoxide has anti-inflammatory effects involving the mitogen-activated protein kinase pathway. Nat Med 2000;6:422e8. Soares MP, Marguti I, Cunha A, et al. Immunoregulatory effects of HO-1: how does it work? Curr Opin Pharmacol 2009;9:482e9. Chen P, Sun B, Chen H, et al. Effects of carbon monoxide releasing moleculeliberated CO on severe acute pancreatitis in rats. Cytokine 2010;49:15e23. Norman JG, Fink GW, Denham W, et al. Tissue-specific cytokine production during experimental acute pancreatitis. A probable mechanism for distant organ dysfunction. Dig Dis Sci 1997;42:1783e8. Norman JG, Fink GW, Franz MG. Acute pancreatitis induces intrapancreatic tumor necrosis factor gene expression. Arch Surg 1995;130:966e70. Norman JG, Fink GW, Messina J, et al. Timing of tumor necrosis factor antagonism is critical in determining outcome in murine lethal acute pancreatitis. Surgery 1996;120:515e21. Sennello JA, Fayad R, Pini M, et al. Interleukin-18, together with interleukin-12, induces severe acute pancreatitis in obese but not in nonobese leptin-deficient mice. Proc Natl Acad Sci U S A 2008;105:8085e90. Pezzilli R, Miniero R, Cappelletti O, et al. Behavior of serum interleukin 12 in human acute pancreatitis. Pancreas 1999;18:247e51. Pereda J, Sabater L, Aparisi L, et al. Interaction between cytokines and oxidative stress in acute pancreatitis. Curr Med Chem 2006;13:2775e87. Norman JG, Fink G, Franz M, et al. Active interleukin-1 receptor required for maximal progression of acute pancreatitis. Ann Surg 1996;223:163e9. Norman JG, Fink GW, Sexton C, et al. Transgenic animals demonstrate a role for the IL-1 receptor in regulating IL-1beta gene expression at steady-state and during the systemic stress induced by acute pancreatitis. J Surg Res 1996;63:231e6. Norman J, Franz M, Messina J, et al. Interleukin-1 receptor antagonist decreases severity of experimental acute pancreatitis. Surgery 1995;117:648e55. Norman J, Yang J, Fink G, et al. Severity and mortality of experimental pancreatitis are dependent on interleukin-1 converting enzyme (ICE). J Interferon Cytokine Res 1997;17:113e18. Paszkowski AS, Rau B, Mayer JM, et al. Therapeutic application of caspase 1/ interleukin-1beta-converting enzyme inhibitor decreases the death rate in severe acute experimental pancreatitis. Ann Surg 2002;235:68e76. Oquendo P, Alberta J, Wen DZ, et al. The platelet-derived growth factor-inducible KC gene encodes a secretory protein related to platelet alpha-granule proteins. J Biol Chem 1989;264:4133e7. Introna M, Bast RC Jr, Tannenbaum CS, et al. The effect of LPS on expression of the early “competence” genes JE and KC in murine peritoneal macrophages. J Immunol 1987;138:3891e6. Bhatia M, Saluja AK, Hofbauer B, et al. The effects of neutrophil depletion on a completely noninvasive model of acute pancreatitis-associated lung injury. Int J Pancreatol 1998;24:77e83.

679

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology

Insulin resistance and low-density apolipoprotein B-associated lipoviral particles in hepatitis C virus genotype 1 infection Simon H Bridge,1 David A Sheridan,1 Daniel J Felmlee,1 Søren U Nielsen,2 Howard C Thomas,3 Simon D Taylor-Robinson,3 R Dermot G Neely,4 Geoffrey L Toms,1 Margaret F Bassendine1 < An additional table is

published online only. To view this file please visit the journal online (http://gut.bmj.com). 1

Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK 2 Institute for Cell and Molecular Biosciences, Newcastle University, Newcastle upon Tyne, UK 3 Division of Diabetes, Endocrinology and Metabolism, Imperial College London, London, UK 4 Department of Clinical Biochemistry, Newcastle upon Tyne Hospitals NHS Foundation Trust, Royal Victoria Infirmary, Newcastle upon Tyne, UK Correspondence to Professor Margaret F Bassendine, Institute of Cellular Medicine, William Leech Building, The Medical School, Newcastle University, Framlington Place, Newcastle upon Tyne NE2 4HH, UK; [email protected] Revised 28 July 2010 Accepted 24 August 2010 Published Online First 12 October 2010

ABSTRACT Background The density of hepatitis C virus (HCV) in plasma is heterogeneous but the factors which influence this are poorly understood. Evidence from animal models and cell culture suggest that low-density apolipoprotein B (apoB)-associated HCV lipoviral particles (LVP) are more infectious than high-density HCV. Objective To measure LVP in patients with chronic hepatitis C genotype 1 (CHC-G1) and examine metabolic determinants of LVP load. Patients 51 patients with CHC-G1 infection. Methods Fasting lipid profiles and homeostasis model assessment of insulin resistance (HOMA-IR) were determined in 51 patients with CHC-G1. LVP and nonLVP viral load were measured by real-time PCR of plasma at density <1.07 g/ml and >1.07 g/ml, respectively, following iodixanol density gradient ultracentrifugation. The LVP ratio was calculated using the formula: LVP/(LVP + non-LVP). Results The mean LVP ratio was 0.241 but varied 25-fold (from 0.029 to 0.74). Univariate analysis showed that the LVP ratio correlated with HOMA-IR (p¼0.004) and the triglyceride/high-density lipoprotein cholesterol (TG/HDLC) ratio (p¼0.004), but not with apoB. In multivariate analysis, HOMA-IR was the main determinant of LVP load (log10IU/ml) (R2¼16.6%; p¼0.037) but the TG/HDL-C ratio was the strongest predictor of the LVP ratio (R2¼24.4%; p¼0.019). Higher LVP ratios were associated with non-response to antiviral therapy (p¼0.037) and with greater liver stiffness (p¼0.001). Conclusion IR and associated dyslipidaemia are the major determinants of low-density apoB-associated LVP in fasting plasma. This provides a possible mechanism to explain why IR is associated with more rapidly progressive liver disease and poorer treatment outcomes.

The characterisation of hepatitis C virus (HCV) particles from the plasma of infected patients shows a highly heterogeneous population of viruses (buoyant density <1.03e1.25 g/ml) due to physical association of virions with host lipoproteins.1e3 HCV particles found in the low-density fractions of plasma are bound to very-low-density lipoprotein (VLDL), intermediate-density lipoprotein (IDL) and low-density lipoprotein (LDL) particles, each of which contains one molecule of the nonexchangeable structural protein, apolipoprotein B (apoB). Low-density virus particles that contain at least apoB, HCV RNA and core proteins have been 680

Significance of this study What is already known about the subject?

< Hepatitis C virus (HCV) RNA-containing particles

circulate in the serum of patients with chronic hepatitis C (CHC) at a range of densities between 1.03 g/ml and 1.25 g/ml. < Experimental data indicate that the low-density HCV particles are infectious in animal models. < In vitro silencing apolipoprotein B (apoB) mRNA leads to a reduction in the secretion of HCV. < Insulin resistance in CHC genotype 1 (CHC-G1) infection predicts faster progression to fibrosis and cirrhosis and a poorer response to antiviral therapy.

What are the new findings? < The proportion of apoB-associated lipoviral

particles (LVP ratio) varies at least 25-fold between patients. < Metabolic determinants of LVP in fasting CHCG1 patients include insulin resistance and a high triglyceride/high-density lipoprotein-cholesterol ratio. < In patients with CHC-G1 a higher LVP ratio is associated with more fibrosis and a poorer response to antiviral therapy.

How might it impact on clinical practice in the forseeable future? < Measurement of low-density apoB-associated LVP provides additional prognostic information and needs to be further evaluated in clinical trials, in particular those using lipid-modulating and insulin-sensitising agents as adjuvant therapies.

termed lipoviral particles (LVP).4 Most LVP from the plasma of an immunodeﬁcient HCV-infected patient without anti-HCV antibodies can be immunoprecipitated with antibodies against apoB.5 LVP may be reliably isolated by density gradient ultracentrifugation in iodixanol which maintains the integrity of the particles.5 This method permits LVP quantitation independent of immune complex formation in immunocompetent patients.4 The development of an efﬁcient HCV cell culture (HCVcc) system has highlighted a critical role of VLDL assembly, maturation and secretion in the HCV life cycle.6 7 In vitro, two types of HCVcc Gut 2011;60:680e687. doi:10.1136/gut.2010.222133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology particles are detected, with a small proportion of low-density particles of high infectivity whereas the majority of released particles lack infectivity.8 Silencing apoB messenger RNA in infected cells causes a 70% reduction in the secretion of both apoB100 and HCV.9 In addition, using HCVcc, Lindenbach et al showed that the virus could establish a long-term infection in chimpanzees and in uPA+/+-SCID mice containing human liver grafts. The virus isolated from infected animals was of lower buoyant density and higher speciﬁc infectivity than HCVcc passaged in cell culture alone, providing further evidence that interaction with host triglyceride-rich lipoproteins plays an important role in the life cycle of HCV.10 This observation was in agreement with other in vivo evidence suggesting that lower density HCV particles are more infectious. When HCV particles derived from human plasma were administered to chimpanzees, a low buoyant density fraction was higher in infectivity than a fraction in which particles were predominantly of high density.11 12 Furthermore, phylogenetic analysis of an acutely infected individual revealed that the infecting viral quasispecies most closely matched quasispecies of low-density virus of the previous host.13 In view of the evident importance of lipoprotein metabolism in the HCV life cycle and the apparent higher infectivity of lowdensity HCV, we used iodixanol density gradient ultracentrifugation and sensitive real-time PCR to quantitate LVP in the fasting plasma of a large cohort of well-characterised patients with chronic hepatitis C genotype 1 (CHC-G1) and examined the host factors and metabolic determinants inﬂuencing LVP load.

(ELISA; Dako UK Ltd, Ely, UK) and assessment of insulin resistance (IR) was performed by calculation of Homeostasis Model Assessments of Insulin Resistance (HOMA-IR) and Quantitative Insulin Sensitivity Check Index (QUICKI) from the serum sample using the formulae: HOMA-IR¼(fasting glucose (mmol/l)3insulin (mU/l)/22.5; QUICKI¼1/(fasting glucose (log10 mmol/l)+fasting insulin (log10 mmol/l)). Apolipoprotein A-I and apoB were measured on each sample by automated rate nephelometric methods (BNII, Dade Behring Marburg GmbH, Marburg, Germany). Liver stiffness was measured within 3 months of the patient’s visit by the non-invasive method of ultrasound transient elastography using a Fibroscan with a standard probe (Echosens, Paris, France).16 HCV genotype was previously determined by the Health Protection Agency North East Laboratories, Newcastle upon Tyne using the line-probe assay (INNOLiPA HCV; Innogenetics, Ghent, Belgium).17

Venous blood and plasma preparation In patients and healthy volunteers, 12 h fasting blood samples were collected in EDTA vacutainer tubes (BD Biosciences Ltd, Oxford, UK) and incubated at 378C. The plasma was separated by centrifugation at 3000 rpm for 10 min at 378C in a Rotanta 460R Hettich centrifuge (DJB Labcare, Buckinghamshire, UK). The blood was handled as close to body temperature as possible to avoid the formation of a cryoglobulin precipitate.18 Complete protease inhibitor cocktail (Roche Products Ltd, Welwyn Garden City, UK) was added to the plasma to prevent protein degradation.

METHODS Study cohort

Iodixanol density gradient ultracentrifugation

Consecutive patients with CHC-G1 infection attending the viral hepatitis clinic at the Freeman Hospital, Newcastle upon Tyne were invited to attend the Clinical Research Facility (CRF), Royal Victoria Inﬁrmary for a fasting blood test and were given a patient information leaﬂet explaining the study. Both treatment-naïve and previous non-responders to combination pegylated interferon-a and ribavirin antiviral therapy were eligible for inclusion. Patients were excluded if they were alcohol dependent, being treated with concurrent lipid-lowering therapy, coinfected with hepatitis B virus, hepatitis delta virus or HIV or had poor venous access due to injecting drug use. Non-responders were also invited to participate in a separate ongoing lipidmodulating intervention study (REC number-07/S0501/21) which, in addition to the above, excluded participants with a body mass index (BMI) $30 kg/m2. Fifty-one patients with CHC-G1 aged >18 years who provided written informed consent attended the CRF after a 12 h fast; 18 were non-responders with a BMI <30 kg/m2. The HCV cohort attending our viral hepatitis clinic comprises approximately 45% patients with CHC-G1, and the study participants were representative of both the regional and national HCV-infected population.14 15

Plasma (0.5 ml) was added to 9.5 ml 12.5% iodixanol solution (2.1 ml iodixanol (60%) (Optiprep, Axis-Shield, Kimbolton, UK), 200 ml 100 mM EDTA, pH 8.0, 200 ml 0.5 M Tris-HCl pH 8.0, 7.15 ml 0.25 M sucrose, 25 ml 2 M MgSO4, 25 ml 2 M MgCl2) in polycarbonate centrifuge tubes (Beckman Coulter, High Wycombe, UK) and inverted several times to mix thoroughly before centrifuging at 50 000 rpm for 24 h at 48C in a type Ti50 rotor and a L8-70M ultracentrifuge (Beckman Coulter). Gradients were harvested according to whether they were from the LVP assay development cohort (seven healthy volunteers and nine patients with CHC) or the validation cohort (51 patients with CHC-G1 infection). For the development cohort the gradients were harvested from the top by collecting 203500 ml fractions and the density of each fraction was measured using a digital refractometer (Atago, Washington, USA). On the basis of the results obtained from the development cohort, the validation cohort gradients were harvested into two fractions: a top 3.5 ml low-density (<1.07 g/ml) fraction (LVP) and a bottom 6.5 ml high-density (>1.07 g/ml) fraction (non-LVP). The densities between these two fractions were measured. All fractions were pulse-vortexed to ensure thorough mixing.

Clinical and laboratory assessment

SDS-PAGE and apoB western blotting

Each subject attending the CRF were reassessed for past alcohol intake and medication history and the following data were collected: sex, age, weight, height, waist and hip circumference. A serum sample was collected for lipid analysis. Total cholesterol, triglyceride (TG), HDL-cholesterol (HDL-C), non-esteriﬁed fatty acids and fasting glucose were measured by standard automated enzymatic methods using an Olympus AU 640 analyser (Olympus, Watford, UK). LDL cholesterol (LDL-C) was calculated using the Friedewald equation and non-HDL cholesterol (non-HDL-C) was obtained by subtraction of HDL-C from total cholesterol. Fasting insulin was measured by immunoassay

Iodixanol fraction samples were prepared by boiling in Laemmli buffer (4% sodium dodecyl sulphate (SDS), 20% glycerol, 10% 2-mercaptoethanol, 0.004% bromophenol blue, 0.125 M Tris HCl) for 10 min. Proteins were separated in 3e18% gradient gels by SDS-polyacrylamide gel electrophoresis (PAGE) on a Bio-Rad Protean II system. ApoB was detected by western blotting after transfer of the resolved proteins to a Hybond polyvinylidene diﬂuoride membrane (GE Healthcare Life Sciences, Little Chalfont, UK) using an enhanced chemiluminescence (ECL) detection kit (GE Healthcare Life Sciences) and polyclonal anti-apoB antibody (Dako).19

Gut 2011;60:680e687. doi:10.1136/gut.2010.222133

681

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology HCV RNA extraction and quantitation by real-time PCR A plasma sample was collected for viral analysis at the same time as the serum samples for lipid analysis. HCV RNA was extracted from whole and fractionated plasma by QIAamp MinElute Virus Spin Kit according to the manufacturer ’s protocol (Qiagen, Crawley, UK). HCV RNA was extracted from 200 ml of EDTA plasma or the iodixanol LVP/non-LVP fraction and eluted into 100 ml of buffer AVE (Qiagen). Extracted HCV RNA was quantitated by two-step real-time PCR for HCV as described previously.19 Reverse transcription was performed using the NCR-3 primer and AMV reverse transcriptase (Promega, Southampton, UK). The HCV positive-strand assay was calibrated against the WHO 3rd international standard for HCV RNA (NIBSC, Potters Bar, UK). Real-time PCR was conducted using an ABI Prism 7000 with Taqman Universal PCR Master Mix (Applied Biosystems, Warrington, UK). Determinations of duplicate tests were averaged. LVP is deﬁned as the HCV RNA measured in the gradient with a density of <1.07 g/ ml and non-LVP is the HCV RNA measured in the gradient with a density of >1.07 g/ml. The LVP ratio was calculated using the formula: LVP/(LVP + non-LVP).

Statistical analyses All statistical analyses were performed using Minitab 15 software (Minitab Inc, State College, USA). The distribution of continuous data was assessed by normality tests. Age, waist circumference, BMI, total cholesterol, LDL-C and non-HDL-C conformed to a normal distribution. The remaining metabolic and viral variables showed positively skewed non-parametric distributions and were therefore either log10-transformed to parametric distributions before hypothesis tests were applied or non-parametric tests were used. Associations between metabolic parameters and viral load, LVP (log10 IU/ml) and LVP ratio were assessed by univariate linear regression analysis. Pearson correlation was used to test for the association between parametric viral and metabolic variables. These analyses yielded regression coefﬁcients (r) and a probability value (p) from testing against the null hypothesis of no association. Spearman rank correlation analysis was used to test for the association between nonparametric viral and metabolic variables. p Values <0.05 were considered to be signiﬁcant. From the univariate analysis, factors that correlated with p<0.05 were used in a stepwise multivariate linear regression analysis to determine which metabolic factors independently predict LVP (log10 IU/ml) and LVP ratio. Those factors that had p<0.05 after multivariate analysis were considered signiﬁcant independent predictors of the variability of LVP (log10 IU/ml) and LVP ratio, with the degree of association indicated by the R2 value. LVP data from patients with CHC-G1 were divided into those with a low LVP ratio (<median) and those with a high LVP ratio ($median). For parametric data, the F test was applied to test the assumption of equal variances and two-sample t tests were used to compare clinical and metabolic parameters between those with a high/ low LVP ratio. For non-parametric data, KruskaleWallis tests were applied to compare groups.The c2 test was used to test the association between high/low LVP ratio with early antiviral treatment responses, metabolic syndrome and liver stiffness.

RESULTS Density distribution of plasma apoB on iodixanol density gradients Plasma from a healthy volunteer was added to a range of isotonic iodixanol gradients with increasing concentrations of iodixanol: 6.25%, 12.5%, 18.75%, 25%, 31.25%, 37.5%, 43.75% 682

and 50% and apoB was detected in the fractionated gradients by SDS-PAGE and western blotting. The optimum concentration of iodixanol was 12.5% as this formed the most linear density gradient. Increasing the concentration of iodixanol reduced the linearity of the density gradient and concentrated the apoBcontaining fractions to the top of the gradient (data not shown).

Development cohort Plasma from seven healthy volunteers and nine patients with CHC were fractionated by iodixanol density ultracentrifugation: 2030.5 ml fractions were collected. ApoB was not detected in the healthy volunteers in fractions with a density >1.062 g/ml and was not detected in patients with CHC with a density >1.07 g/ml (see table 1 in online supplement). This density cutoff value in a 12.5% iodixanol gradient was subsequently used to deﬁne LVP in our assay.

Validation cohort Iodixanol gradients were fractionated into the LVP fraction at a density of <1.07 g/ml (top 3.5 ml of the gradient) and the nonLVP fraction at a density of >1.07 g/ml (remaining 6.5 ml of the gradient).

Clinical and metabolic characteristics The physical and metabolic characteristics of the 51 patients with CHC-G1 infection are shown in table 1. Forty-eight patients were Caucasian and the remaining three patients were Chinese (n¼1), sub-Saharan African (n¼1) and Eastern Mediterranean/Middle Eastern (n¼1). BMI was normal (<25 kg/m2) in 24 patients (47%), overweight (25e30 kg/m2) in 20 (39%) and obese (>30 kg/m2) in 7 (14%) patients. According to the criteria of the International Diabetes Federation, 12 patients had metabolic syndrome.21 Of these, two patients were receiving treatment for type 2 diabetes mellitus. Table 1 Clinical and metabolic characteristics of patients with hepatitis C virus genotype 1 (HCV-G1) infection Clinical and metabolic characteristics

HCV-G1 (n[51)

Sex (M/F) Age (years) Liver stiffness $13 kPa, n (%)* Alanine transaminase (U/l) Waist (cm) Waist/hip ratio Body mass index (kg/m2) Total cholesterol (mmol/l) LDL cholesterol (mmol/l) Non-HDL cholesterol (mmol/l) Triglyceride (mmol/l) Apolipoprotein B (g/l) HDL cholesterol (mmol/l) Apolipoprotein A1 (g/l) Triglyceride/HDL cholesterol Non-esterified fatty acids (mmol/l) Fasting glucose (mmol/l) Fasting insulin (mU/l) HOMA-IRy QUICKIz

37/14 48.0610.4 14 (29) 92.0658.0 89612 0.9660.06 25.464.1 4.5960.99 2.7360.86 3.3660.95 1.3560.76 0.90 60.25 1.2360.32 1.4260.28 1.2360.91 0.4460.30 4.9760.65 7.9464.15 1.8261.09 1.2160.39

Data shown are mean6SD unless otherwise indicated. *Liver stiffness on ultrasound transient elastography as measured by Fibroscan with a standard probe.16 A score $13.0 kPa represents cirhosis.20 yAssessment of insulin resistance was performed by calculation of Homeostasis Model Assessments of Insulin Resistance (HOMA-IR) from a fasting blood sample using the formula: (fasting glucose (mmol/l)3 insulin (mU/l))/22.5. zQUICKI¼1/(fasting glucose (log10mmol/l) + fasting insulin (log10mmol/l)).

Gut 2011;60:680e687. doi:10.1136/gut.2010.222133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Metabolic determinants of LVP Univariate analysis showed that LVP load (log10 IU/ml) in fasted patients correlated positively with glucose (table 2; r¼0.307, p¼0.028), insulin (table 2; r¼0.393, p¼0.005) and HOMA-IR (table 2; r¼0.383, p¼0.005, ﬁgure 2B). Univariate analysis showed that the LVP ratio correlated positively with glucose (table 2; r¼0.311, p¼0.026), insulin (r¼0.393, p¼0.005) and HOMA-IR (ﬁgure 2C; r¼0.397, p¼0.004). The LVP ratio also correlated positively with TG levels (ﬁgure 3A; r¼0.320, p¼0.022), the TG/HDL-C ratio (ﬁgure 3B; r¼0.392, p¼0.004) and with liver stiffness (table 2; r¼0.419, p¼0.003). Multivariate

Figure 1 Population distribution of viral parameters. (A) Total plasma hepatitis C virus (HCV) RNA viral load (log10 IU/ml). (B) Lipoviral particle (LVP) load (log10 IU/ml) found at a density of <1.07 g/ml. (C) LVP ratio (fraction of LVP load to LVP + non-LVP load (log10 IU/ml)).

Viral parameters Viral load, LVP and non-LVP load were measured in all 51 patients with CHC-G1 infection. The results of viral load and LVP load are summarised in ﬁgure 1A,B. Viral load and LVP load produced similar distributions. The majority (65%) of viral loads were >5.90 log10 IU/ml. The LVP ratio was calculated and the results are shown in ﬁgure 1C. The LVP ratio varied widely between patients ranging from 0.029 to 0.74 (mean 0.241; median 0.177).

Insulin resistance As expected,22 HOMA-IR was strongly correlated with waist circumference (r¼0.641, p<0.0001), BMI (r¼0.526, p<0.0001) and positively correlated with age (r¼0.324, p¼0.020). There was no correlation between HOMA-IR and HCV viral load (ﬁgure 2A, r¼0.078, p¼0.585). Liver stiffness correlated with HOMA-IR (r¼0.381, p¼0.007). Gut 2011;60:680e687. doi:10.1136/gut.2010.222133

Figure 2 Relationship between total viral load, lipoviral particle (LVP) load and LVP ratio with HOMA-IR. Open circles indicate non-responders, closed circles indicate those with an early virological response (EVR), open triangles indicate those not treated. Correlation between HOMA-IR and (A) viral load (log10 IU/ml) (r¼0.078; p¼0.585), (B) LVP load (log10IU/ml) (r¼0.383; p¼0.005) and (C) LVP ratio (r¼0.397; p¼0.004). 683

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Table 2

Univariate analysis of viral load and LVP in CHC-G1 with host and metabolic parameters in patients with CHC-G1

Host and metobolic characteristics

Viral load (log10 IU/ml)

LVP (log10 IU/ml)

LVP ratio

Age (years)* Waist* Body mass index (kg/m2)* Fibroscan measurement of liver stiffnessy Total cholesterol (mmol/l)* Non-HDL cholesterol (mmol/l)* Triglyceride (mmol/l)y Apolipoprotein B (g/l)y HDL cholesterol (mmol/l)y Apolipoprotein A1 (g/l)y Triglyceride/HDL cholesteroly Non-esterified fatty acids (mmol/l)y Fasting glucose (mmol/l)y Fasting insulin (mU/l)y HOMA-IRy QUICKIy

r¼0.229; p¼0.106 r¼0.141; p¼0.333 r¼0.159; p¼0.265 r¼0.086; p¼0.555 r¼0.181; p¼0.203 r¼0.185; p¼0.194 r¼0.101; p¼0.479 r¼0.178; p¼0.216 r¼0.178; p¼0.216 r¼-0.118; p¼0.416 r¼0.082; p¼0.568 r¼0.082; p¼0.568 r¼0.171; p¼0.231 r¼0.121; p¼0.407 r¼0.078; p¼0.585 r¼0.121; p¼0.414

r¼0.163; p¼0.253 r¼0.210; p¼0.147 r¼0.249; p¼0.078 r¼0.124; p¼0.395 r¼0.018; p¼0.899 r¼0.058; p¼0.697 r¼0.248; p¼0.079 r¼0.014; p¼0.921 r¼0.014; p¼0.921 r¼0.136; p¼0.347 r¼0.256; p¼0.070 r¼0.109; p¼454 r[0.307; p[0.028 r[0.393; p[0.028 r[0.383; p[0.005 r[L0.387; p[0.007

r¼0.043; p¼0.765 r¼0.158; p¼0.278 r¼0.077; p¼0.592 r[0.419; p[0.003 r¼0.141; p¼0.884 r¼0.021; p¼0.884 r[0.320; p[0.022 r¼0.073; p¼0.615 r¼0.073; p¼0.615 r¼0.257; p¼0.072 r[0.392; p[0.004 r¼0.128; p¼0.381 r[0.311; p[0.026 r[0.393; p[0.005 r[0.397; p[0.004 r[L0.371; p[0.009

*Pearson’s correlation analysis. ySpearman rank correlation analysis. CHC-G1, chronic hepatitis C genotype 1; HDL, high-density lipoprotein; HOMA-IR, Homeostasis Model Assessments of Insulin Resistance; LVP, lipoviral particle; QUICKI, Quantitative Insulin Sensitivity Check Index.

regression analysis showed that HOMA-IR was the only signiﬁcant independent determinant of LVP load (log10 IU/ml) (R2¼16.6%; p¼0.037), whereas TG/HDL-C ratio was the only independent predictor of LVP ratio (R2¼24.4%; p¼0.019). The comparison of clinical and metabolic variables in patients with CHC-G1 with a low LVP ratio (deﬁned as below the median value of 0.177; n¼25) and those with a high LVP ratio (deﬁned as above the median value of 0.177; n¼26) is shown in table 3. Patients with a low LVP ratio had higher LDL-C (p¼0.037) while those with a high LVP ratio had a signiﬁcantly higher fasting plasma glucose (p¼0.044), insulin (p¼0.005) and HOMA-IR (p¼0.008). The latter patients also had a higher TG level (p¼0.015), lower HDL-C (p¼0.015) and hence higher TG/ HDL-C ratio (p¼0.003). In addition, patients with CHC-G1 with a high LVP ratio had a larger waist circumference (p¼0.037) and greater liver stiffness (p¼0.001). Further analysis of the LVP ratio shows that patients with a high LVP ratio were more likely to have a liver stiffness measurement of >13 kPa (c2¼5.95; p¼0.015), consistent with advanced ﬁbrosis.

maintains the integrity of these particles. This allows the use of sensitive real-time PCR methods for accurate determination of HCV RNA associated with low-density (<1.07 g/ml) apoBcontaining lipoproteins (LVP load) and the proportion of total HCV RNA found in plasma as LVP (LVP ratio). Our study shows

LVP and treatment response Data on the early virological response (EVR) to antiviral treatment with pegylated interferon-a and ribavirin was available for 42 patients with CHC-G1. Of these, 18 (43%) were nonresponders with the remaining 24 (57%) having achieved a complete or partial EVR (HCV RNA negative or >2 log reduction in HCV RNA at 12 weeks). Eleven patients on treatment had metabolic syndrome and 10 of these patients were nonresponders. The median LVP ratio in the non-responder group was signiﬁcantly higher than in the EVR group (0.338 vs 0.201, p¼0.031, ﬁgure 4). Conversely, patients with CHC-G1 with a low LVP ratio were more likely to have an EVR (p¼0.037, table 3).

DISCUSSION This study shows that IR is an independent predictor of LVP load and that the TG/HDL-C ratio is an independent predictor of the LVP ratio in patients with CHC-G1 infection. The LVP ratio is a relative measure of the density distribution of HCV in the plasma. This is the ﬁrst study to measure LVP in a wellcharacterised cohort of patients with CHC using a method based on fractionation of iodixanol density gradients which 684

Figure 3 Relationship between lipoviral particle (LVP) ratio with triglyceride levels and the triglyceride/HDL-C ratio. Open circles indicate non-responders, closed circles indicate those with an early virological response (EVR), open triangles indicate those not treated. Correlations between LVP ratio with (A) triglyceride (r¼0.320; p¼0.022) and (B) triglyceride/HDL-C (r¼0.392; p¼0.004). Gut 2011;60:680e687. doi:10.1136/gut.2010.222133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Table 3

Relationship between LVP ratio with metabolic variables, liver stiffness and EVR to antiviral therapy

Metabolic variable

Low LVP ratio (n[25)

High LVP ratio (n[26)

p Value*

Age (years)y Body mass index (kg/m2)y Waist circumference (cm)y Liver stiffness (kPa)y Total cholesterol (mmol/l)y LDL cholesterol (mmol/l) Non-HDL cholesterol (mmol/l)y Triglyceride (mmol/l)z Apolipoprotein B (g/l)z HDL-C (mmol/l)z Apolipoprotein A (g/l)z Triglyceride/HDL-Cz Glucose (mmol/l)z Insulin (mU/l)z HOMA-IRz EVR, n (% of total treated)x { Non-response, n (% of total treated)x { Fibroscan score >13 kPa (n)x Fibroscan score <13 kPa (n)x

46.6611.6 2564.1 85.7611.1 9.53 4.8560.98 2.9960.8 3.4960.91 1.0860.46 0.9360.25 1.3660.33 1.560.29 0.8460.4 4.8460.65 5.9662.75 1.2560.68 13 (31) 4 (9.5) 3 21

49.369 25.964.2 92.7612.2 19.54 4.3460.95 2.4960.86 3.2360.98 1.6260.89 0.8760.25 1.1160.25 1.3560.25 1.6261.1 5.160.64 9.5664.67 2.1761.3 11 (26.2) 14 (33.3) 11 14

0.359 0.443 0.037 0.001 0.065 0.037 0.331 0.015 0.317 0.015 0.065 0.003 0.044 0.005 0.008 c2¼4.36; p¼0.037

c2¼5.954; p¼0.015

Data shown are mean6SD unless otherwise indicated. *LVP ratio: high versus low. yParametric t test. zNon-parametric KruskalleWallis test. xc2 test for a 2 3 2 table. {42 patients were treated with antiviral therapy. EVR, early virological response; HDL, high-density lipoprotein; HOMA-IR, Homeostasis Model Assessments of Insulin Resistance; LDL, low-density lipoprotein; LVP, lipoviral particle.

that the LVP ratio varied widely in fasting patients, ranging from 0.029 to 0.740 and that, on average, LVP contribute approximately a quarter (w24%) of total viral load. Using univariate analysis, we found a positive correlation of LVP load with HOMA-IR. The LVP ratio correlated signiﬁcantly with HOMAIR, TG/HDL-C ratio and TG levels, but there was no association between total viral loads and host clinical or metabolic parameters. Using multivariate analysis, we showed that HOMA-IR accounts for approximately 16% of the variability of the LVP load. Interestingly, in the subgroup of 42 patients with CHC-G1 in the present study for whom data on response to antiviral treatment was available, the LVP ratio was signiﬁcantly higher in non-responders than in those with a complete or partial EVR, consistent with the LVP fraction contributing to persistent

Figure 4 Relationship between lipoviral particle (LVP) ratio with early antiviral treatment outcome. LVP ratio and early antiviral treatment outcomes (EVR to non-responders, p¼0.031). The box height represents the interquartile range (Q1eQ3), the line within the box is the median value, the lower whisker represents Q1e1.5 (Q3eQ1) and the upper whisker represents Q3+1.5 (Q3eQ1). Gut 2011;60:680e687. doi:10.1136/gut.2010.222133

infection. Furthermore, patients with CHC-G1 with a higher LVP ratio had more advanced ﬁbrosis (liver stiffness >13 kPa). IR in CHC is clinically relevant because it not only predicts faster progression to cirrhosis but also a poorer response to antiviral therapy. In subjects with CHC-G1, IR and overt diabetes mellitus are major determinants of advanced ﬁbrosis.23e27 IR is also a risk factor for progression of ﬁbrosis in HCV-infected patients following liver transplantation.28 IR is an independent predictor of a sustained virological response (SVR) in patients with CHC receiving pegylated interferon-a/ribavirin.22 29e32 Our results show that IR is the most important determinant of absolute LVP load in CHC-G1 infection, while the strongest predictor of LVP ratio was found to be the TG/HDL-C ratio which is characteristically increased in IR. IR and TG levels have previously been associated with high viral load in some large cohort studies. In a study of 500 patients with CHC, those with IR had a high serum viral load (deﬁned as a serum HCV RNA level >600 000 IU/ml) more frequently than those without IR (55.3% vs 42.3%, p<0.009).33 In another study of >500 patients with CHC, both HOMA-IR and TG levels were associated with higher HCV viral loads (p<0.05).34 In a recent study of >200 patients with CHC in a single US centre where patients with co-existing metabolic syndrome are more common, hypertriglyceridaemia correlated with higher viral loads (p¼0.042).35 Epidemiological data indicate a strong risk for the development of IR and, ultimately, type 2 diabetes mellitus in patients with CHC infection.36e38 In a recent 5-year prospective followup study, HCV infection per se and genotype 1 were independent risk factors for the development of IR.39 IR in CHC is most strongly associated with genotypes 1 and 4 infection.33 HCV may promote IR through genotype-speciﬁc molecular mechanisms,40 41 although recent evidence suggests that IR in CHC is mainly in muscle and not liver.42 Furthermore, recent studies demonstrate that effective clearance of HCV with antiviral therapy correlates with an improvement in IR and reduced incidence of onset of type 2 diabetes mellitus.43 44 685

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology In normal individuals without CHC the liver secretes 1018 VLDL particles daily, with smaller VLDL2 being constitutively secreted and production of larger TG-rich VLDL1 being regulated by insulin in normoglycaemic subjects.45 46 IR results in overproduction of large low-density TG-rich VLDL1 particles.47 48 Kinetic studies have shown that <20% of large TG-rich VLDL1 are converted to LDL; most are cleared directly from the plasma.49 50 The positive correlation of the LVP ratio with the TG/HDL-C ratio, even in the fasting state, implies that HCV preferentially associates with triglyceride-rich lipoproteins which, in the fasting state, will predominantly be large VLDLdthat is, VLDL1, not chylomicrons. Fatty acids for the biosynthesis of VLDL1 are derived from TG stored in lipid droplets,51 and HCV uses lipid droplets for the production of infectious virions.52 We have also recently shown in vivo that very-low-density HCV (<1.025 g/ml) surges 26-fold after a high fat meal.53 Collectively, our in vivo studies are consistent with LVP in the plasma preferentially associating with large TG-rich VLDL. Association between HCV and VLDL1 as LVP may enable these putative infectious particles to escape antibody neutralisation54 and be rapidly cleared by the liver via the apoE remnant pathway.55 Our data suggest that the measurement of LVP and calculation of the LVP ratio is clinically important and provides additional information when measured in combination with conventional HCV viral load. For example, incorporation of LVP measurement in therapeutic trials using insulin-sensitising or lipid-modulating drugs as adjuvant therapies31 56e58 may help identify patients who may beneﬁt from this approach. The development of a simple and inexpensive method to measure LVP would facilitate the analysis of LVP in appropriatelypowered larger prospective studies and would provide conﬁrmation as to whether LVP is clinically useful in predicting patient responsiveness to antiviral regimens. In summary, we report the ﬁrst accurate measurements of LVP fraction of total hepatitis C viral load in a well-characterised cohort of patients with CHC-G1. Our results show that the LVP ratio varies at least 25-fold between patients and is signiﬁcantly associated with metabolic parameters and important clinical outcomes. In particular, a higher LVP ratio is strongly associated with IR and higher TG levels, both components of the metabolic syndrome. Higher LVP ratios were also found in non-responders to antiviral treatment and those with more advanced ﬁbrosis. Our in vivo study offers further insight into the life cycle of HCV and suggests that HCV preferentially associates with TGrich VLDL1 which is modulated by IR. Our data support the hypothesis that promotion of IR by HCV infection may drive the production of more LVP, providing a possible explanation of why patients with CHC-G1 with IR have a worse prognosis. Acknowledgements The authors thank Fiona I Fenwick for her smooth running of the laboratories and are grateful to the hepatitis C patients for their generous donation of time and inconvenience. Funding Supported by the Medical Research Council (G0502028) and the Newcastle upon Tyne Healthcare Charity. The funders had no role in the study design, data collection and analysis. HCT and SDT-R were supported by the NIHR Biomedical Facility at Imperial College London. Competing interests None. Ethics approval This study was conducted with the approval of the Northumberland Research ethics committee (REC number-07/H0902/45).

2. 3.

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

Provenance and peer review Not commissioned; externally peer reviewed. 31.

REFERENCES 1.

686

Watson JP, Bevitt DJ, Spickett GP, et al. Hepatitis C virus density heterogeneity and viral titre in acute and chronic infection: a comparison of immunodeficient and immunocompetent patients. J Hepatol 1996;25:599e607.

32.

Prince AM, Huima-Byron T, Parker TS, et al. Visualization of hepatitis C virions and putative defective interfering particles isolated from low-density lipoproteins. J Viral Hepat 1996;3:11e17. Pumeechockchai W, Bevitt D, Agarwal K, et al. Hepatitis C virus particles of different density in the blood of chronically infected immunocompetent and immunodeficient patients: Implications for virus clearance by antibody. J Med Virol 2002;68:335e42. Andre P, Komurian-Pradel F, Deforges S, et al. Characterization of low- and verylow-density hepatitis C virus RNA-containing particles. J Virol 2002;76:6919e28. Nielsen SU, Bassendine MF, Burt AD, et al. Association between hepatitis C virus and very-low-density lipoprotein (VLDL)/LDL analyzed in iodixanol density gradients. J Virol 2006;80:2418e28. Gastaminza P, Kapadia SB, Chisari FV. Differential biophysical properties of infectious intracellular and secreted hepatitis C virus particles. J Virol 2006;80:11074e81. Huang H, Sun F, Owen DM, et al. Hepatitis C virus production by human hepatocytes dependent on assembly and secretion of very low-density lipoproteins. Proc Natl Acad Sci U S A 2007;104:5848e53. Miyanari Y, Atsuzawa K, Usuda N, et al. The lipid droplet is an important organelle for hepatitis C virus production. Nat Cell Biol 2007;9:1089e97. Nahmias Y, Goldwasser J, Casali M, et al. Apolipoprotein B-dependent hepatitis C virus secretion is inhibited by the grapefruit flavonoid naringenin. Hepatology 2008;47:1437e45. Lindenbach BD, Meuleman P, Ploss A, et al. Cell culture-grown hepatitis C virus is infectious in vivo and can be recultured in vitro. Proc Natl Acad Sci U S A 2006;103:3805e9. Bradley D, McCaustland K, Krawczynski K, et al. Hepatitis C virus: buoyant density of the factor VIII-derived isolate in sucrose. J Med Virol 1991;34:206e8. Hijikata M, Shimizu YK, Kato H, et al. Equilibrium centrifugation studies of hepatitis C virus: evidence for circulating immune complexes. J Virol 1993;67:1953e8. Diaz O, Cubero M, Trabaud MA, et al. Transmission of low-density hepatitis C viral particles during sexually transmitted acute resolving infection. J Med Virol 2008;80:242e6. Sheridan DA, Price DA, Schmid ML, et al. Apolipoprotein B-associated cholesterol is a determinant of treatment outcome in patients with chronic hepatitis C virus infection receiving anti-viral agents interferon-alpha and ribavirin. Aliment Pharmacol Ther 2009;29:1282e90. Costella A, Irvine N, McCartney M, et al. Hepatitis C in the UK 2009. London: Health Protection Agency Centre for Infections, 2009. Ziol M, Handra-Luca A, Kettaneh A, et al. Noninvasive assessment of liver fibrosis by measurement of stiffness in patients with chronic hepatitis C. Hepatology 2005;41:48e54. Stuyver L, Wyseur A, van Arnhem W, et al. Second-generation line probe assay for hepatitis C virus genotyping. J Clin Microbiol 1996;34:2259e66. Agnello V, Elfahal M. Cryoglobulin types and rheumatoid factors associated with clinical manifestations in patients with hepatitis C virus infection. Dig Liver Dis 2007;39(Suppl 1):S25e31. Nielsen SU, Bassendine MF, Burt AD, et al. Characterization of the genome and structural proteins of hepatitis C virus resolved from infected human liver. J Gen Virol 2004;85:1497e507. Friedrich-Rust M, Ong MF, Martens S, et al. Performance of transient elastography for the staging of liver fibrosis: a meta-analysis. Gastroenterology 2008;134:960e74. Alberti KG, Zimmet P, Shaw J. The metabolic syndromeea new worldwide definition. Lancet 2005;366:1059e62. Romero-Gomez M, Del Mar Viloria M, Andrade RJ, et al. Insulin resistance impairs sustained response rate to peginterferon plus ribavirin in chronic hepatitis C patients. Gastroenterology 2005;128:636e41. Fartoux L, Poujol-Robert A, Guechot J, et al. Insulin resistance is a cause of steatosis and fibrosis progression in chronic hepatitis C. Gut 2005;54:1003e8. Muzzi A, Leandro G, Rubbia-Brandt L, et al. Insulin resistance is associated with liver fibrosis in non-diabetic chronic hepatitis C patients. J Hepatol 2005;42:41e6. Petta S, Camma C, Di Marco V, et al. Insulin resistance and diabetes increase fibrosis in the liver of patients with genotype 1 HCV infection. Am J Gastroenterol 2008;103:1136e44. Merchante N, Rivero A, de Los Santos-Gil I, et al. Insulin resistance is associated with liver stiffness in HIV/HCV co-infected patients. Gut 2009;58:1654e60. Svegliati-Baroni G, Bugianesi E, Bouserhal T, et al. Post-load insulin resistance is an independent predictor of hepatic fibrosis in virus C chronic hepatitis and in nonalcoholic fatty liver disease. Gut 2007;56:1296e301. Veldt BJ, Poterucha JJ, Watt KD, et al. Insulin resistance, serum adipokines and risk of fibrosis progression in patients transplanted for hepatitis C. Am J Transplant 2009;9:1406e13. Poustchi H, Negro F, Hui J, et al. Insulin resistance and response to therapy in patients infected with chronic hepatitis C virus genotypes 2 and 3. J Hepatol 2008;48:28e34. Chu CJ, Lee SD, Hung TH, et al. Insulin resistance is a major determinant of sustained virological response in genotype 1 chronic hepatitis C patients receiving peginterferon alpha-2b plus ribavirin. Aliment Pharmacol Ther 2009;29:46e54. Khattab M, Emad M, Abdelaleem A, et al. Pioglitazone improves virological response to peginterferon alpha-2b/ribavirin combination therapy in hepatitis C genotype 4 patients with insulin resistance. Liver Int 2010;30:447e54. Mizuta T, Kawaguchi Y, Eguchi Y, et al. Whole-body insulin sensitivity index is a highly specific predictive marker for virological response to peginterferon plus

Gut 2011;60:680e687. doi:10.1136/gut.2010.222133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology

33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.

ribavirin therapy in chronic hepatitis C patients with genotype 1b and high viral load. Dig Dis Sci 2010;55:183e9. Moucari R, Asselah T, Cazals-Hatem D, et al. Insulin resistance in chronic hepatitis C: association with genotypes 1 and 4, serum HCV RNA level, and liver fibrosis. Gastroenterology 2008;134:416e23. Hsu CS, Liu CH, Liu CJ, et al. Factors affecting early viral load decline of Asian chronic hepatitis C patients receiving pegylated interferon plus ribavirin therapy. Antivir Ther 2009;14:45e54. Forde KA, Law C, O’Flynn R, et al. Do statins reduce hepatitis C RNA titers during routine clinical use? World J Gastroenterol 2009;15:5020e7. Mehta SH, Brancati FL, Strathdee SA, et al. Hepatitis C virus infection and incident type 2 diabetes. Hepatology 2003;38:50e6. Huang JF, Dai CY, Hwang SJ, et al. Hepatitis C viremia increases the association with type 2 diabetes mellitus in a hepatitis B and C endemic area: an epidemiological link with virological implication. Am J Gastroenterol 2007;102:1237e43. Serfaty L, Capeau J. Hepatitis C, insulin resistance and diabetes: clinical and pathogenic data. Liver Int 2009;29(Suppl 2):13e25. Park SK, Cho YK, Park JH, et al. The change of insulin sensitivity in hepatitis C patients with normal insulin sensitivity: a 5-year prospective follow-up study. Intern Med J 2010;40:503e11. Pazienza V, Clement S, Pugnale P, et al. The hepatitis C virus core protein of genotypes 3a and 1b downregulates insulin receptor substrate 1 through genotypespecific mechanisms. Hepatology 2007;45:1164e71. Douglas MW, George J. Molecular mechanisms of insulin resistance in chronic hepatitis C. World J Gastroenterol 2009;15:4356e64. Milner KL, van der Poorten D, Trenell M, et al. Chronic hepatitis C is associated with peripheral rather than hepatic insulin resistance. Gastroenterology 2010;138:932e41.e1e3. Arase Y, Suzuki F, Suzuki Y, et al. Sustained virological response reduces incidence of onset of type 2 diabetes in chronic hepatitis C. Hepatology 2009;49:739e44. Delgado-Borrego A, Jordan SH, Negre B, et al. Reduction of insulin resistance with effective clearance of hepatitis C infection: results from the HALT-C trial. Clin Gastroenterol Hepatol 2010;8:458e62. Gill JM, Brown JC, Bedford D, et al. Hepatic production of VLDL1 but not VLDL2 is related to insulin resistance in normoglycaemic middle-aged subjects. Atherosclerosis 2004;176:49e56.

46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57.

58.

Adiels M, Taskinen MR, Packard C, et al. Overproduction of large VLDL particles is driven by increased liver fat content in man. Diabetologia 2006;49:755e65. Adiels M, Olofsson SO, Taskinen MR, et al. Diabetic dyslipidaemia. Curr Opin Lipidol 2006;17:238e46. Adiels M, Olofsson SO, Taskinen MR, et al. Overproduction of very low-density lipoproteins is the hallmark of the dyslipidemia in the metabolic syndrome. Arterioscler Thromb Vasc Biol 2008;28:1225e36. Packard CJ, Munro A, Lorimer AR, et al. Metabolism of apolipoprotein B in large triglyceride-rich very low density lipoproteins of normal and hypertriglyceridemic subjects. J Clin Invest 1984;74:2178e92. Demant T, Packard C. In vivo studies of VLDL metabolism and LDL heterogeneity. Eur Heart J 1998;19(Suppl H):H7e10. Olofsson SO, Bostrom P, Andersson L, et al. Lipid droplets as dynamic organelles connecting storage and efflux of lipids. Biochim Biophys Acta 2009;1791:448e58. McLauchlan J. Lipid droplets and hepatitis C virus infection. Biochim Biophys Acta 2009;1791:552e9. Felmlee DJ, Sheridan DA, Bridge SH, et al. Intravascular Transfer Contributes to Postprandial Increase in Numbers of Very-Low-Density Hepatitis C Virus Particles. Gastroenterology 2010;139:1774e83, 1783.e1e6. Grove J, Nielsen S, Zhong J, et al. Identification of a residue in hepatitis C virus E2 glycoprotein that determines scavenger receptor BI and CD81 receptor dependency and sensitivity to neutralizing antibodies. J Virol 2008;82:12020e9. Agnello V, Abel G, Elfahal M, et al. Hepatitis C virus and other flaviviridae viruses enter cells via low density lipoprotein receptor. Proc Natl Acad Sci U S A 1999;96:12766e71. Romero-Gomez M, Diago M, Andrade RJ, et al. Treatment of insulin resistance with metformin in naive genotype 1 chronic hepatitis C patients receiving peginterferon alfa-2a plus ribavirin. Hepatology 2009;50:1702e8. Vierling JM, Hamzeh FM, Lentz EL, et al. Virologic and metabolic responses in chronic hepatitis C (CHC) patients with insulin resistance (IR) treated with pioglitazone and peginterferon alpha-2a plus ribavirin. Hepatology 2009; S4:1028A. Harrison SA, Rossaro L, Hu KQ, et al. Relationship of the use of statins and elevated low-density lipoprotein (LDL) or total cholesterol (TC) to virologic response in patients treated for hepatitis C virus (HCV) in the ideal study. Hepatology 2009;50:85A.

Editor’s quiz: GI snapshot ANSWER From the question on page 630 The patient had inadvertently swallowed a toothpick which had travelled through the gut until it had perforated the wall of the upper rectum. CT revealed the vertically oriented toothpick which had led to a small inﬂammatory osteolysis of the sacrum (ﬁgure 1). As a complication, a liver abscess had developed causing the inﬂammatory signs. The toothpick could be extracted at sigmoidoscopy. Forty millilitres of pus were aspirated from the abscess. With antibiotic treatment the laboratory abnormalities including thrombocytopaenia gradually returned to normal and the abdominal pain subsided. Five months after discharge the patient is well. Swallowing toothpicks appears to be a transcultural custom.1e5 Most toothpicks already perforate in the gastroduodenal area and necessitate surgery. One wonders how many toothpicks are successful in completing the journey through the gastrointestinal tract. Our toothpick only narrowly missed this goal. Competing interests None. Patient consent Obtained.

Figure 1 Sagittal reconstruction of the CT scans of the pelvis showing the toothpick.

Provenance and peer review Not commissioned; externally peer reviewed.

2.

Gut 2011;60:687. doi:10.1136/gut.2009.190546a

3. 4.

REFERENCES 1.

Erichsen R, Sommer T. Toothpick perforation of the intestine causing abdominal wall abscess. Ugeskr Laeger 2009;171:1609e10.

Gut May 2011 Vol 60 No 5

5.

El-Tarchichi MA, Yafi MF, Debek AH. Electronic clinical challenges and images in GI. Perforating toothpick mimicking Crohn’s disease of the ileum. Gastroenterology 2009;136:e8e9. Liu HJ, Liang CH, Huang B, et al. Migration of a swallowed toothpick into the liver: the value of multiplanar CT. Br J Radiol 2009;82:e79e81. Dente M, Santi F, Solinas L, et al. Laparoscopic diagnosis and management of jejunal perforation resulting from accidental toothpick ingestion. Am Surg 2009;75:178e9. Lanthaler M, Grissmann T, Schwentner L, et al. Unusual differential diagnosis of upper abdominal pain. Diagn Ther Endosc 2009;2009:817052.

687

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology

Community and personal risk factors for hepatitis C virus infection: a survey of 23 820 residents in Taiwan in 1991e2 Mei-Hsuan Lee,1 Hwai-I Yang,1 Chin-Lan Jen,1 Sheng-Nan Lu,2 Shiou-Hwei Yeh,3 Chun-Jen Liu,4 San-Lin You,1 Chien-An Sun,5 Li-Yu Wang,6 Wei J Chen,7 Chien-Jen Chen,1,7 for the R.E.V.E.A.L.eHCV Study Group* 1

Genomics Research Center, Academia Sinica, Taipei, Taiwan 2 Department of Gastroenterology, Kaohsiung Chang-Gung Memorial Hospital, Kaohsiung, Taiwan 3 Department of Microbiology, National Taiwan University, Taipei, Taiwan 4 Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan 5 School of Public Health, Fu-Jen Catholic University, Taipei, Taiwan 6 MacKay Medical College, Taipei, Taiwan 7 Graduate Institute of Epidemiology, College of Public Health, National Taiwan University, Taipei, Taiwan Correspondence to Professor Chien-Jen Chen, Genomics Research Center, Academia Sinica, 128 Academia Road Section 2, Nankang, Taipei 11529, Taiwan; [email protected] *Other members of the Risk Evaluation of Viral Load Elevation and Associated Liver Disease/CancereHepatitis C Virus (REVEALeHCV) study are listed in the appendix. Revised 5 October 2010 Accepted 7 October 2010 Published Online First 10 November 2010

ABSTRACT Aim The aim of this study was to explore the communitylevel risk factors, such as high hepatitis C viruse (HCV)RNA positive rate and limited medical resources in a township, for HCV infection, one major cause of liver cirrhosis and hepatocellular carcinoma. Methods This study enrolled 23 820 residents living in 155 villages of seven townships in Taiwan in 1991e2 to explore both individual and community risk factors for HCV infection. Antibodies against HCV (anti-HCV), HCVRNA and HCV genotype in serum samples were determined by enzyme immunoassay, PCR and melting curve analysis, respectively. Results The overall anti-HCV seroprevalence was 5.5%, HCV-RNA was detectable in 68.1% of the seropositives of anti-HCV, and genotype 1 was the most prevalent genotype (54.6%). Personal risk factors for the seropositivity of anti-HCV included older age, female gender, low educational level and history of blood transfusion. Based on the multilevel analysis, persons living in villages with high HCV-RNA-positive rates and limited healthcare resources had an increased seroprevalence of anti-HCV after adjustment for individual risk factors. The multivariate-adjusted prevalence OR (95% CI) was 3.49 (1.80 to 6.76) and 8.48 (5.07 to 14.20) for villages with medium and high HCV-RNA positive rate, respectively. The multivariateadjusted OR (95% CI) was was 1.75 (0.76 to 4.01) and 3.91 (2.25 to 6.80), respectively, for villages with medium and poor healthcare resources. Conclusions This study suggests that community risk factors contribute significantly to the variation in anti-HCV seroprevalence. It implies both the adequacy of healthcare resources and the treatment of patients positive for HCV-RNA may prevent individual residents from the acquisition of HCV infection from the community. Hepatitis C virus (HCV) infects approximately 130 million people worldwide, roughly 3% of the human population.1 Chronic hepatitis C can progress to end-stage liver diseases2 and accounts for approximately one-third of liver cancers globally,3 presenting a signiﬁcant public health burden. Although hepatitis B virus (HBV) infection contributes most of hepatocellular carcinomas in developing countries in Asia,4 HCV infection is becoming increasingly important as a result of the successful HBV vaccination programme.5 6 Epidemiological characteristics of HCV infection are important for the prevention of its spread in the

688

Significance of this study What is already known about this subject?

< There were great geographical variations in the

seroprevalence of HCV, which could not be accounted for sufficiently by personal risk factors. < Residents living in the same area may share a similar likelihood of acquiring infections and the risk factors for infection with HCV may be classified as individual and community risk factors. < Individual risk factors for HCV infection have been extensively investigated, whereas community risk factors have been rarely evaluated.

What are the new findings?

< A multilevel approach was utilised wherein

individual and community risk factors were simultaneously added to the model to understand which individual or community-level risk factors were associated with personal HCV infection. < Communities with a high HCV-RNA-positive rate or limited medical resources had increased antiHCV seroprevalence after taking into consideration personal risk factors.

How might it impact on clinical practice in the foreseeable future? < Serum HCV-RNA test is important to identify anti-HCV seropositives who may need appropriate interventions to reduce the risk of HCV infection in areas with high HCV-RNA seropositive rates. < Assessment of community-level risk factors enables implementation of a comprehensive community programme for HCV prevention and intervention.

community. However, most previous studies mainly enrolled blood donors or clinical patients rather than the general population as study participants.7e9 Drug users sharing injection equipment is a major risk factor for HCV infection in western countries,10e12 whereas unsafe medical injections using incompletely sterilised equipment are predominant in Taiwan.13 14 Other risk factors

Gut 2011;60:688e694. doi:10.1136/gut.2010.220889

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology included age,15 educational level16 17 and residential area.12 14 16 18 The geographical variability in antibodies against HCV (antiHCV) seroprevalence may be explained by the compositional differences in the risk factors in various regions.19 However, several studies found that the geographical variation still existed after adjustment for individual-level risk factors.14 20 These results suggest personal characteristics alone may be insufﬁcient to account for the variations in anti-HCV seroprevalence among regions, and the identiﬁcation of community-level risk factors may help explain the geographical discrepancy and prevent the spread of HCV in communities. The community-level factors contributing to geographical variations in anti-HCV seroprevalence may include the HCVRNA-positive rate and adequacy of healthcare resources. HCVRNA was detectable in sera samples of 52e80% seropositives of anti-HCV,15 16 20e23 and individuals with elevated serum HCVRNA levels had an increased risk of hepatocellular carcinoma.24 As patients seropositive for anti-HCV and detectable for HCVRNA had a high transmissibility, and residents living in areas with high HCV-RNA-positive rates may have an increased risk of being seropositive for anti-HCV. Individuals received therapeutic injections given by non-licensed healthcare providers were reported to have an elevated risk of anti-HCV seropositivity.13 14 25 Therapeutic equipment might be reused in areas where healthcare resources are relatively inadequate. People living in such areas may thus have a similar likelihood of contracting infection as a result of sharing the same medical facilities. Therefore, inadequate healthcare resources may be a community-level risk factor associated with HCV infection. In our previous caseecontrol study of 272 anti-HCV-seropositive cases and 282 matched seronegative controls,14 we evaluated individual risk factors associated with the HCV infection. Comprehensive information on correlates of antiHCV seropositivity, including a detailed history of blood transfusion, haemodialysis, tattooing and medical injection, was collected through standardised questionnaire interviews. Iatrogenic risk factors such as blood transfusion, medical injection and accupunture as well as tattooing were all signiﬁcantly associated with HCV infection. Among them, a history of blood transfusion was the most important risk factor for HCV infection. Intravenous drug abuse was not a risk factor for HCV infection in the general population of Taiwan. The speciﬁc aim of this study was to assess the importance of township-level risk factors for HCV infection by utilising a multilevel approach.26 The identiﬁcation and intervention of both individual and community-level risk factors may help the prevention of HCV transmission in communities.

MATERIALS AND METHODS Study area and participant enrolment The study area and participant enrolment have been described previously.27 In this community-based cancer screening programme started in 1991e2, seven townships were selected as study areas. They included two northern townships (Sanchi and Chutung) and two southern townships (Potzu and Kaohsu) on the main Taiwan island, and three townships (Makung, Huhsi, and Paihsa) on Penghu islets. There were 12e34 villages in these townships with a total of 155 villages. At study entry, the names, national identiﬁcation numbers and addresses of residents aged 30e65 years were abstracted from the household registration ofﬁces in seven study townships. In total, 89 293 ethnic Chinese individuals (47 079 men and 42 214 women) residing in the seven townships were invited Gut 2011;60:688e694. doi:10.1136/gut.2010.220889

to participate in the study. Invitation letters were initially mailed. Residents who did not respond to the invitation were called on the telephone. Among all invited residents, 23 820 (11 973 men and 11 847 women) provided informed consent and were enrolled in this study. Demographic data for residents who did not participate in this study were quite similar to those of residents who agreed to participate.27 The participants were personally interviewed using structured questionnaires, and received health examinations including biochemical and serological tests at recruitment. Blood samples were obtained using disposable needles and heparinised vacuum syringes. The study protocol was reviewed and approved by the Institutional Review Board of the College of Public Health at National Taiwan University.

Community-level risk factors Participants residing in villages of a study township shared the same living environment including healthcare facilities. Therefore this study explored community-level risk factors including HCV-RNA-positive rates and healthcare resources in each township. The HCV-RNA-positive rate was calculated by the percentage of persons detectable for HCV-RNA among participants seropositive for anti-HCV in each township. It was categorised as low, medium and high HCV-RNA-positive rate for the ﬁrst, second and third, and fourth quartile of HCV-RNA-positive rates. For healthcare resources, number of physicians per 100 000 population was used as a surrogate,28 which was obtained from the government statistical reports in 1991. The healthcare resource index was categorised as poor, medium and adequate by the ﬁrst, second and third, and fourth quartile of the mean number of licenced physicians per 100 000 population in the year 1991 in Taiwan.

Individual-level risk factors Personal interviews by trained public health nurses using a structured questionnaire were carried out to collect information on sociodemographic characteristics, familial history of cancers and lifestyle variables from each participant. Variables including age, sex, educational level and history of blood transfusion were regarded as potential individual risk factors for HCV infection in this study.

Laboratory examinations Blood samples were collected without anticoagulant and were separated by centrifugation at room temperature. Serum samples were stored at 708C. These procedures were completed within 6 h of sample collection. All samples were collected and frozen according to standardised procedures and tested in a central laboratory. Serum samples were tested for anti-HCV using an enzyme immunoassay with commercial kits (Abbott HCV EIA version 2.0; Abbott Laboratories, Illinois, USA; or AxSYM HCV 3.0; Abbott Laboratories). HCV-RNA was measured for samples positive for anti-HCV using the COBAS TaqMan HCV test, v2.0 (Roche Diagnostics, Indianapolis, New Jersey, USA), an in-vitro nucleic acid ampliﬁcation test for the quantiﬁcation of HCV-RNA. The detection limit of serum HCVRNA was 25 IU/ml. Participants with serum HCV-RNA levels equal to or higher than 25 IU/ml were deﬁned as those with detectable HCV-RNA. Serum samples positive for HCV-RNA were examined for HCV genotypes by LightCycler-based PCR and melting curve analysis, which could effectively differentiate different HCV genotypes by showing different melting temperatures. In this test, HCV genotype 1 and HCV genotype non-1 have been examined and the genotype calling for 689

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology genotype 1 and non-1 can achieve a 100% accuracy by the current design.29 30

Statistical analysis Possible trends in age-speciﬁc anti-HCV seroprevalence were evaluated using the CochraneArmitage trend test. The agee sex-standardised anti-HCV seroprevalence was derived for comparison at the village level (using the entire group of study participants as the standard population). Logistic regression models were performed to evaluate potential correlates of detectable HCV-RNA and HCV genotypes in participants seropositive for anti-HCV. All statistical procedures were conducted using SAS software (version 9.1). The data in this study were treated as a two-level hierarchical structure with participants nested within each village. The villages were treated as clusters in the multilevel analysis. The procedures of the multilevel models suggested by Merlo and colleagues31e33 were used in this study. First, an empty model without explanatory variables was used to assess variations of anti-HCV seroprevalence among villages and to test the statistical signiﬁcance of the variation. Then, the individual risk factors were included to investigate the extent of the variation among villages that could be explained by the individual compositional differences. Finally, the model was expanded to incorporate community-level risk factors of villages. The proportional change in variance was used to examine the changes in village-level variances between two different models by including individual risk factors or community-level risk factors.33 The median OR was used to measure village heterogeneity in anti-HCV seroprevalence.31 34 A large median OR (far greater than one) indicated the existence of geographical clustering in anti-HCV seroprevalence. The MLwiN software version 2.035 36 was used for all multilevel analyses.

RESULTS Among 23 820 participants, 35 had inadequate serum samples for the serological test of anti-HCV. The overall seroprevalence of anti-HCV was 5.5% in 23 785 participants. There was a signiﬁcant increasing trend of anti-HCV seroprevalence with increasing age for both men and women. Women had a higher prevalence of HCV infection than men in each age stratum, as shown in table 1. There was a great variation in anti-HCV seroprevalence ranging from 2.0% to 26.4% among seven townships, as shown in table 2. Table 2 also shows the prevalence of detectable HCV-RNA and HCV genotypes in seven townships. Among 1195 participants seropositive for anti-HCV with adequate serum samples for HCV-RNA testing, 789 (68.1%) were positive for HCV-RNA. In general, study townships with a higher anti-HCV seroprevalence had a higher positive rate of HCV-RNA. Among participants seropositive for

Table 1

anti-HCV, men were more likely to be HCV-RNA detectable than women with an age-adjusted OR (95% CI) of 2.24 (1.73 to 2.92). Individuals who had a history of blood transfusions had a 1.41 (95%CI, 1.03 to 1.93) times risk of being positive for HCV-RNA after adjustment for age and sex (data not shown). In table 3 is shown the distribution of HCV genotypes among participants with detectable HCV-RNA. There were 394 (54.6%) individuals infected with HCV genotype 1 and 328 (45.4%) infected with HCV genotype non-1. The frequency distribution of HCV genotype was signiﬁcantly different among four townships in the main Taiwan island (p<0.01), whereas it was similar among three townships in Penghu Islets. Age, sex and blood transfusion history were not signiﬁcantly associated with HCV genotype examined by logistic regressions in this study. There were several villages within the seven townships we studied and the ageesex standardised seroprevalence of antiHCV showed a great variation in these villages. The seroprevalence ranged from 3.3% to 19.9% in 13 villages in Sanchi Township, 1.5e44.6% in 27 villages in Potzu Township, 1.8e16.1% in 25 villages in Chutung Townships, 0e5.9% in 19 villages in Kaohsu Township, 0.6e50.1% in 15 villages in Paihsa Township, 0e35.1% in 22 villages in Huhsi Township and 0e6.9% in 34 villages in Makung Township. The variation at the village level was more striking than that at the township level. The number of villages by different anti-HCV seroprevalence within seven townships is displayed in table 4 to show the variation in anti-HCV seroprevalence among villages. Table 5 shows the measures of association with individual and community-level risk factors as well as measures of variation or clustering for seroprevalence of anti-HCV derived from multilevel logistic regression models. In the empty model (model 0 median OR (95% credible interval) of 2.95 (2.56 to 3.48), suggesting that the heterogeneity of anti-HCV seroprevalence among villages was noteworthy. The variance was almost not reduced (only 1.6%) when individual characteristics for anti-HCV seropositivity were included into the model (model 1). Older ages, low educational levels and a history of blood transfusion were signiﬁcantly associated with an increased seroprevalence of anti-HCV. When community-level risk factors, that is, the adequacy of healthcare resources and HCV-RNA seropositive rate in villages were included, the village-level variance decreased 57.4%, compared with the variance in the model with individual risk factors alone (model 1). The results suggested that the inclusion of community-level risk factors was important. In the ﬁnal model (model 2), two community-level risk factors and four individual-level risk factors were included to evaluate their associations with anti-HCV seroprevalence. Among participants living in the same village, older ages, low educational levels and a history of blood transfusion remained signiﬁcantly associated with an increased risk of anti-HCV seropositivity.

Seroprevalence of antibodies against HCV by age and sex in 1991e2 in Taiwan Total population

Men

Women

Age (years)*

No tested

No of positive

Prevalence with 95% CI (%)

No tested

No of positive

Prevalence with 95% CI (%)

No tested

No of positive

Prevalence with 95% CI (%)

30e39 40e49 50e59 60e65 Overall

6820 6240 7295 3430 23785

226 297 529 261 1313

3.3 4.8 7.3 7.6 5.5

3263 2943 3695 2067 11968

107 120 223 127 577

3.3 4.1 6.0 6.1 4.8

3557 3297 3600 1363 11817

119 177 306 134 736

3.4 5.4 8.5 9.8 6.2

(2.9 (4.2 (6.7 (6.7 (5.2

to to to to to

3.7) 5.3) 7.9) 8.5) 5.8)

(2.7 (3.4 (5.3 (5.1 (4.4

to to to to to

3.9) 4.8) 6.8) 7.2) 5.2)

(2.8 (4.6 (7.6 (8.3 (5.8

to to to to to

3.9) 6.1) 9.4) 11.4) 6.7)

*p for trend <0.01 for both men and women. HCV, hepatitis C virus.

690

Gut 2011;60:688e694. doi:10.1136/gut.2010.220889

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Table 2 Taiwan

Prevalence of antibodies against HCV (anti-HCV) and prevalence HCV-RNA viraemia in 1991e2, stratified by seven study townships in Anti-HCV

Townships

No tested

Taiwan main island Sanchi Potzu Chutung Kaohsu Penghu islands Paihsa Huhsi Makung Total population

HCV-RNAy No of positive

Prevalence with 95% CI (%)

No tested*

No of positive

Prevalence with 95% CI (%)

1450 3600 5402 3874

206 252 220 77

14.2 7.0 4.1 2.0

(12.4 to 16.0) (6.2 to 7.8) (3.5 to 4.6) (1.6 to 2.4)

182 198 194 69

130 148 127 40

71.4 74.8 65.5 58.0

(64.9 (68.7 (58.8 (46.3

to to to to

78.0) 80.8) 72.2) 69.6)

1305 1840 6314 23785

345 71 142 1313

26.4 3.9 2.3 5.5

(24.1 to 28.8) (3.0 to 4.7) (1.9 to 2.6) (5.2 to 5.8)

318 68 130 1159

229 38 77 789

72.0 55.9 59.2 68.1

(67.1 (44.1 (50.8 (65.4

to to to to

76.9) 67.7) 67.7) 70.8)

*One hundred and fifty-four anti-hepatitis C virus (HCV)-positive serum samples were inadequate for the HCV-RNA test. yPrevalence of HCV-RNA viraemia among participants seropositive for anti-HCV.

Participants living in a village with a high HCV-RNA-positive rate had an increased risk of being seropositive for anti-HCV compared with those living in a village with a low HCV-RNApositive rate. The odds ratios were 3.49 (1.80e6.76) and 8.48 (5.07e14.20), respectively, for participants living in villages with medium and high positive rates in comparison with those living in villages with low HCV-RNA-positive rates. Participants living in a village with relatively limited healthcare resources also had an elevated risk of HCV infection than those living in a village with adequate healthcare resources. Compared with those living in villages with adequate medical resources as the reference group, participants living in villages with medium and poor medical resources had 1.75 (0.76e4.01) and 3.91 (2.25e6.80) times the risk of being seropositive for anti-HCV.

DISCUSSION This community-based study in Taiwan found an anti-HCV seroprevalence of 5.5%, which was moderately high compared with the seroprevalence ranging from 0.9% to 12.6% in other large surveys.15 16 21 22 There are six genotypes of HCV,37 which may be used to determine the duration and dosage of interferon therapy and to predict the antiviral response.38 HCV genotype 1 was reported as the most prevalent genotype with a proportion ranging from 51% to 74% in community-based studies in western countries.16 21e23 Previous studies found signiﬁcant geographical variations in the distribution of HCV genotypes.8 9 21 However, HCV genotypes could only be determined among anti-HCV seropositives with detectable HCV-RNA, and the genotype-

Table 3

speciﬁc HCV prevalence might be imprecisely estimated by low HCV infection rates.8 21 A large number of participants from the general population are essential for the accurate assessment of geographical variation in HCV genotypes. In this study, the proportions of HCV genotype 1 and non-1 were quite different in four study townships on the main Taiwan island, but similar in three townships in Penghu Islets. It might be interpreted that these three townships in Penghu Islets are located closely to each other and shared common healthcare facilities as also observed in Italy.39 Women in this study had a higher anti-HCV seroprevalence than men, whereas men in western countries had a higher antiHCV seroprevalence than women.10 11 Disposable medical equipment was not commonly used throughout Taiwan before 1980, when the glucose-based nutrient or vitamin injections were frequently prescribed to sick people. As women were more concerned about their health status or minor illnesses, they were more likely to receive the injection than men. Men infected with HCV might also have a higher mortality rate than infected women, and anti-HCV prevalence in men would thus be more likely to be lower than women due to a faster attrition of the infected. Obstetric measures might also put women at higher risk than men for HCV infection. Unfortunately, we did not collect data on parity from woman participants to examine this hypothesis. The increasing anti-HCV seroprevalence with advanced age observed in this study was similar to those found in Japan and India.15 19 Older participants may have more chances to receive multiple injections, which resulted in a higher cumulative HCV infection risk during their lifetime. Furthermore, older

Prevalence of HCV genotypes among participants with detecf HCV-RNA in 1991e2, stratified by seven study townships in Taiwan Genotype 1

Townships Taiwan main island Sanchi Potzu Chutung Kaohsu Penghu islands Paihsa Huhsi Makung Total population

No tested

No

Genotype non-1 Prevalence with 95% CI (%)

No

Prevalence with 95% CI (%)

117 139 119 36

92 88 36 9

78.6 63.3 30.3 25.0

(71.2 (55.3 (22.0 (10.9

to to to to

86.1) 71.3) 38.5) 39.2)

25 51 83 27

21.4 36.7 69.8 75.0

(13.9 (28.7 (61.5 (60.9

to to to to

28.8) 44.7) 78.0) 89.2)

210 31 70 722

115 15 39 394

54.8 48.4 55.7 54.6

(48.0 (30.8 (44.1 (50.9

to to to to

61.5) 66.0) 67.4) 58.2)

95 16 31 328

45.2 51.6 44.3 45.4

(38.5 (34.0 (32.7 (41.8

to to to to

52.0) 69.2) 55.9) 49.1)

HCV, hepatitis C virus.

Gut 2011;60:688e694. doi:10.1136/gut.2010.220889

691

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Table 4 Number of villages by various age-sex standardised anti-HCV seroprevalence within seven townships in Taiwan in 1991e2 Age-sex standardised anti-HCV seroprevalence (%) Township

No of villages

<2.5

2.5e<5

5e<10

10e<20

‡20

Sanchi Chutung Potzu Kaohsu Makung Huhsi Paihsa

13 25 27 19 34 22 15

0 7 1 13 24 8 1

2 11 9 15 8 8 0

0 4 13 1 2 2 1

9 2 3 0 0 0 1

2 1 1 0 0 2 12

HCV, hepatitis C virus.

participants who lived through World War II (1939e45) may have an increased risk of exposure to HCV infection due to the shortage of healthcare resources. The individual risk of anti-HCV seropositivity may be personally and ecologically determined. Individual risk factors for HCV infection have been extensively investigated.14e17 40 However, what community-level risk factors determine antiHCV seropositivity was rarely investigated. To achieve this goal, a multilevel analysis has been adopted in a growing number of recent epidemiological studies,26 because it provides more accurate standard error estimations by accounting for the nonindependence of observations nested within contexts. There were two studies using multilevel analysis to examine the risk factors of HCV infection.41 42 These two studies were limited to families of study subjects with speciﬁc exposure. In our study, only 1.6% of geographical variation in anti-HCV seroprevalence was attributable to individual risk factors. In comparison, 57.4% of geographical variation in anti-HCV seroprevalence was due to ecological factors at the village level. The ﬁnding suggests the importance of looking beyond individual-level risk factors into community-level risk factors in the study of HCV infection. A previous ecological study found residents living in less afﬂuent areas had an increased prevalence of anti-HCV,18 which

supported our ﬁndings that healthcare resources in villages were important in the determination of geographical variation in antiHCV seroprevalence. The application of mapping technology in this study is useful for summarising and displaying information in a contextually and spatially meaningful fashion. Unfortunately, no information on age, gender, education level and socioeconomic status was collected for each individual in each study area. As HCV infection may be more prevalent in some demographic groups in some study areas, the lack of individual risk factors might result in difﬁculties in the interpretion of the ﬁndings and ecological fallacy (the associations between afﬂuence and HCV prevalence at the community level might not be found at the individual level). Healthcare providers might reuse syringes or needles when delivering medications in order to save money in villages with limited resources. There might also be some non-licensed health workers who provided unsafe injections without using standard procedures in villages with poorer healthcare resources. Unsafe injections might be the most important transmission route for HCV in these villages. A recent study found that persons born in high HCV endemic area had an increased risk of being positive for anti-HCV,12 suggesting that HCV infection may be determined by ecological characteristics. An important ecological factor that determines anti-HCV seropositivity was the HCV-RNA-positive rate in communities. This study found that participants living in villages with high HCV-RNA detectable rates had an increased seroprevalence of anti-HCV. Viraemic participants positive for both anti-HCV and HCV-RNA have highly replicating HCV, which may be the major transmission source in villages. Previous studies found infected mothers with detectable HCV-RNA had an increased likelihood to transmit the virus to their infants.43 Relatives living in families of anti-HCV-seropositive patients with detectable HCV-RNA had a twofold risk of HCV infection compared with those living in families of patients negative for HCV-RNA.41 Our ﬁndings imply that anti-HCV seropositives with undetectable HCV-RNA were those who cleared the virus and they were no longer a source of efﬁcient transmission.

Table 5 Measures of association with individual and community-level risk factors as well as measures of variation or clustering for seroprevalence of anti-HCV derived from multilevel logistic regression models Model 0 (empty model) Measures of association (OR, 95% CI) Individual-level risk factors Sex (male vs female) Age (in 10-year increment) Educational level (vs 7+ years) 1e6 years Illiteracy History of blood transfusion (yes vs no) Community-level risk factors HCV-RNA-positive rate (vs low)* Medium High Adequacy of healthcare resource (vs adequate)y Medium Poor Measures of variation or clustering Village-level variance (SE) PCV MOR (95% CrI)

Model 1

Model 2

0.92 (0.80 to 0.99) 1.20 (1.12 to 1.30)

0.92 (0.81 to 1.01) 1.19 (1.12 to 1.27)

1.69 (1.44 to 1.99) 2.02 (1.63 to 2.45) 3.55 (3.04 to 4.14)

1.65 (1.42 to 1.94) 1.92 (1.58 to 2.33) 3.26 (2.81 to 3.77)

3.49 (1.80 to 6.76) 8.48 (5.07 to 14.20) 1.75 (0.76 to 4.01) 3.91 (2.25 to 6.80) 1.29 (0.18) 2.95 (2.56 to 3.48)

1.27 (0.18) 1.6% 2.93 (2.54 to 3.45)

0.54 (0.11) 57.4% 2.06 (1.83 to 2.35)

*Categorised as high, medium and low for the first, second and third, and fourth quartile of hepatitis C virus (HCV)-RNA-positive rate among anti-HCV seropositives in townships. yLicensed physician number per 100 000 population, categorised as adequate, medium and poor for the first, second and third, and fourth quartile of number of physicians per 100 000 population in Taiwan in 1991. CrI, credible interval; MOR median odds ratio PCV proportional change in variance.

692

Gut 2011;60:688e694. doi:10.1136/gut.2010.220889

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Whereas participants with detectable HCV-RNA had active HCV replication and played as infection sources by iatrogenic transmission routes. It highlighted the importance of the serum HCV-RNA test for patients who are seropositive for anti-HCV to identify those with persistent HCV infection in order to control the spread of the virus. The ﬁnding of a trend of increasing anti-HCV prevalence with the increase in the HCV-RNA seropositive rate suggested that individuals with detectable serum HCV-RNA levels may be the major infection source of HCV in the community. Anti-HCV seropositives identiﬁed from community surveillance should be referred for further examination of the serum HCV-RNA level and necessary treatment to prevent the spread of HCV in the community. The trend of increasing anti-HCV prevalence with the decreased adequacy of healthcare resources suggests the importance of education on HCV prevention provided to healthcare workers in areas with inadequate resources. To our knowledge, this study is the ﬁrst to incorporate individual and community-level risk factors to elucidate the geographical variation of anti-HCV seroprevalence. In this community-based cancer screening programme, we screened three important cancers in Taiwan, namely liver cancer, cervical cancer and nasopharyngeal cancer. As HBV and HCV are the most important causes of liver cancer in Taiwan, we tested seromarkers including serum levels of alanine aminotransferase and alpha-fetoproten and serostatus of hepatitis B surface antigen (HBsAg) and anti-HCV to identify high-risk participants for regular examinations by ultrasonography. Interestingly, among 1313 anti-HCV-seropositive participants, 218 were seropositive for HBsAg showing a dual infection rate of 0.92% (218/23820) in the entire study participants. The HBsAg-seropositive prevalence was 14.2% (112/789) and 19.5% (72/370), respectively, for those with and without detectable serum HCVRNA levels, suggesting that HBV might be suppressed in those with active HCV infection. In this study we did not collect detailed information on risk factors associated with HCV infection for the entire cohort. However, iatrogenic risk factors were found to be signiﬁcant correlates for HCV infection in our previous caseecontrol study.14 The past or current drug abuse was not collected because illicit drug use was infrequent in Taiwan,44 45 particularly in our agriculture/ﬁshery-based study townships. It is criminal to sell or use illicit drug in Taiwan. Although injection drug use may contribute signiﬁcantly to HCV infection in western countries, it was considered a less important risk factor as found in our previous studies. The strengths of this community-based study were the enrolment of a large group of participants and the application of a multilevel analysis to identify both individual and community risk factors associated with HCV infection. The large sample size ensured the precision of seroprevalence estimates, and enabled the detection of geographical variation. By means of multilevel analysis, community-level risk factors associated with anti-HCV seroprevalence were assessed. It would be valuable to focus interventions on speciﬁc ecological factors in order to reduce individual risk of HCV infection. In addition to allocating more heathcare resources in inadequate villages, the treatment of patients with persistent HCV infection will reduce the risk of HCV infection in areas with high HCV-RNA-positive rates. Assessment of community-level risk factors will enable us to implement a comprehensive community programme for HCV prevention and intervention. In conclusion, people living in the same villages share a similar risk of being infected with HCV. This cross-sectional survey conducted in 1991e2 indicated that community-level risk Gut 2011;60:688e694. doi:10.1136/gut.2010.220889

factors such as high HCV-RNA-positive rates and shortage of healthcare resources played important roles in the geographical variation in anti-HCV seroprevalence. The detection and treatment of patients with detectable HCV-RNA and avoidance of unsafe injections are essential for the prevention of new HCV infections. Funding This study was supported by research grants from Department of Health, Executive Yuan, Taipei, Taiwan; Bristol-Myers Squibb Co, USA; Academia Sinica, Taipei, Taiwan; and National Health Research Institutes (NHRI-EX98-9806PI), Chunan, Taiwan. Competing interests None. Patient consent Obtained. Ethics approval This study was conducted with the approval of the institutional review board of the College of Public Health at National Taiwan University. Contributors CJC had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: MHL, CJC. Acquisition of data: HIY, CLJ, SLY, CJC. Analysis and interpretation of data: MHL, HIY, WJC, CJC. Drafting of the manuscript: MHL. Critical revision of the manuscript for important intellectual content: MHL, HIY, CLJ, SHY, CLL, SLY, CAS, LYW, WJC, CJC. Statistical analysis: MHL. Obtained funding: CJC. Administrative, technical, or material support: HIY, CLJ, SNL, SHY, CJL, SLY, CAS, LYW, WJC, CJC. Study supervision: CJC. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13. 14. 15. 16. 17. 18.

Hutin Y, Kitler ME, Dore GJ, et al. Global burden of disease (GBD) for hepatitis C. J Clin Pharmacol 2004;44:20e9. Fattovich G, Stroffolini T, Zagni I, et al. Hepatocellular carcinoma in cirrhosis: incidence and risk factors. Gastroenterology 2004;127:S35eS50. Parkin DM. The global health burden of infection-associated cancers in the year 2002. Int J Cancer 2006;118:3030e44. Raza SA, Clifford GM, Franceschi S. Worldwide variation in the relative importance of hepatitis B and hepatitis C viruses in hepatocellular carcinoma: a systematic review. Br J Cancer 2007;96:1127e34. Chien YC, Jan CF, Kuo HS, et al. Nationwide hepatitis B vaccination program in Taiwan: effectiveness in the 20 years after it was launched. Epidemiol Rev 2006;28:126e35. Lu SN, Su WW, Yang SS, et al. Secular trends and geographic variations of hepatitis B virus and hepatitis C virus-associated hepatocellular carcinoma in Taiwan. Int J Cancer 2006;119:1946e52. Shepard CW, Finelli L, Alter MJ. Global epidemiology of hepatitis C virus infection. Lancet Infect Dis 2005;5:558e67. Dusheiko G, Schmilovitzweiss H, Brown D, et al. Hepatitis C virus genotypes an investigation of type-specific differences in geographic origin and disease. Hepatology 1994;19:13e18. Martinot-Peignoux M, Roudot-Thoraval F, Mendel I, et al. Hepatitis C virus genotypes in France: relationship with epidemiology, pathogenicity and response to interferon therapy. J Viral Hepat 1999;6:435e43. Armstrong GL, Wasley A, Simard EP, et al. The prevalence of hepatitis C virus infection in the United States, 1999 through 2002. Ann Intern Med 2006;144:705e14. Perez CM, Suarez E, Torres EA, et al. Seroprevalence of hepatitis C virus and associated risk behaviours: a population-based study in San Juan, Puerto Rico. Int J Epidemiol 2005;34:593e9. Meffre C, Le Strat Y, Delarocque-Astagneau E, et al. Prevalence of hepatitis B and hepatitis C virus infections in France in 2004: social factors are important predictors after adjusting for known risk factors. J Med Virol 2010;82:546e55. Wang CS, Chang TT, Chou P. Differences in risk factors for being either a hepatitis B carrier or anti-hepatitis C+ in a hepatoma-hyperendemic area in rural Taiwan. J Clin Epidemiol 1998;51:733e8. Sun CA, Chen HC, Lu CF, et al. Transmission of hepatitis C virus in Taiwan: prevalence and risk factors based on a nationwide survey. J Med Virol 1999;59:290e6. Chowdhury A, Santra A, Chaudhuri S, et al. Hepatitis C virus infection in the general population: a community-based study in West Bengal, India. Hepatology 2003;37:802e9. Alter MJ, Kruszon-Moran D, Nainan OV, et al. The prevalence of hepatitis C virus infection in the United States, 1988 through 1994 [see comment]. N Engl J Med 1999;341:556e62. Chang HC, Yu MW, Lu CF, et al. Risk factors associated with hepatitis C virus infection in Taiwanese government employees. Epidemiol Infect 2001;126:291e9. Mujeeb SA, Shahab S, Hyder AA. Geographical display of health information: study of hepatitis C infection in Karachi, Pakistan [erratum appears in Public Health 2001 Jan; 115(1):82]. Public Health 2000;114:413e15.

693

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

33.

Wasley A, Alter MJ. Epidemiology of hepatitis C: geographic differences and temporal trends. Semin Liver Dis 2000;20:1e16. Ansaldi F, Bruzzone B, Salmaso S, et al. Different seroprevalence and molecular epidemiology patterns of hepatitis c virus infection in Italy. J Med Virol 2005;76:327e32. Dubois F, Desenclos JC, Mariotte N, et al. Hepatitis C in a French population-based survey, 1994: seroprevalence, frequency of viremia, genotype distribution, and risk factors. Hepatology 1997;25:1490e6. Guadagnino V, Stroffolini T, Rapicetta M, et al. Prevalence, risk factors, and genotype distribution of hepatitis C virus infection in the general population: a community-based survey in southern Italy. Hepatology 1997;26:1006e11. McMahon BJ, Hennessy TW, Christensen C, et al. Epidemiology and risk factors for hepatitis C in Alaska natives. Hepatology 2004;39:325e32. Lee MH, Yang HI, Lu SN, et al. Hepatitis C virus seromarkers and subsequent risk of hepatocellular carcinoma: long-term predictors from a community-based cohort study. J Clin Oncol 2010;28:4587e93. Sun CA, Chen HC, Lu SN, et al. Persistent hyperendemicity of hepatitis C virus infection in Taiwan: the important role of iatrogenic risk factors. J Med Virol 2001;65:30e4. Diez-Roux AV. Multilevel analysis in public health research. Annu Rev Public Health 2000;21:171e92. Iloeje UH, Yang HI, Jen CL, et al. Risk and predictors of mortality associated with chronic hepatitis B infection. Clin Gastro Hepatol 2007;5:921e31. Armstrong D, Barnett E, Casper M, et al. Community occupational structure, medical and economic resources, and coronary mortality among US blacks and whites, 1980e1988. Ann Epidemiol 1998;8:184e91. Liu CJ, Chuang WL, Lee CM, et al. Peginterferon alfa-2a plus ribavirin for the treatment of dual chronic infection with hepatitis B and C viruses. Gastroenterology 2009;136:496e504. Yeh SH, Tsai CY, Kao JH, et al. Quantification and genotyping of hepatitis B virus in a single reaction by real-time PCR and melting curve analysis. J Hepatol 2004;41:659e66. Merlo J, Chaix B, Ohlsson H, et al. A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena. J Epidemiol Community Health 2006;60:290e7. Merlo J, Chaix B, Yang M, et al. A brief conceptual tutorial on multilevel analysis in social epidemiology: interpreting neighbourhood differences and the effect of neighbourhood characteristics on individual health. J Epidemiol Community Health 2005;59:1022e8. Merlo J, Yang M, Chaix B, et al. A brief conceptual tutorial on multilevel analysis in social epidemiology: investigating contextual phenomena in different groups of people. J Epidemiol Community Health 2005;59:729e36.

34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.

Larsen K, Merlo J. Appropriate assessment of neighborhood effects on individual health: integrating random and fixed effects in multilevel logistic regression [see comment]. Am J Epidemiol 2005;161:81e8. Browne W. MCMC estimation in MLwiN version 20. London, UK: Centre for Multilevel Modelling, Institute of Education, University of London, 2003;297. Rasbash J, Steele F, Browne W. A user’s guide to MLwiN, version 20 documentation vVersion 21e. London: Centre for Multilevel Modelling, Institute of Education, University of London, 2003. Simmonds P. Viral heterogeneity of the hepatitis C virus. J Hepatol 1999;31:54e60. Scott JD, Gretch DR. Molecular diagnostics of hepatitis C virus infectionea systematic review. JAMA 2007;297:724e32. Dal Molin G, Ansaldi F, Biagi C, et al. Changing molecular epidemiology of hepatitis C virus infection in northeast Italy. J Med Virol 2002;68:352e6. Shin HR, Kim JY, Kim JI, et al. Hepatitis B and C virus prevalence in a rural area of South Korea: the role of acupuncture. Br J Cancer 2002;87:314e18. Akhtar S, Moatter T. Multilevel modeling of intra-household spread of hepatitis C virus infection, Karachi, Pakistan. Am J Trop Med Hyg 2007;76:446e9. McMahon JM, Pouget ER, Tortu S. Individual and couple-level risk factors for hepatitis C infection among heterosexual drug users: a multilevel dyadic analysis. J Infect Dis 2007;195:1572e81. Tovo PA, Pembrey L, Newell ML, et al. A significant sex-but not elective cesarean sectioneeffect on mother-to-child transmission of hepatitis C virus infection. J Infect Dis 2005;192:1872e9. Chen KT, Chen CJ, Fagot-Campagna A, et al. Tobacco, betel quid, alcohol, and illicit drug use among 13- to 35-year-olds in I-Lan, rural Taiwan: prevalence and risk factors. Am J Public Health 2001;91:1130e4. Chiang SC, Chen CY, Chang YY, et al. Prevalence of heroin and methamphetamine male users in the northern Taiwan, 1999e2002: capture-recapture estimates. BMC Public Health 2007;7.

APPENDIX Other members of the REVEALeHCV study group: National Taiwan University Hospital: CY Hsieh, HS Lee, PM Yang, CH Chen, JD Chen, SP Huang, CF Jan; National Taiwan University: THH Chen; National Defence Medical Center: CA Sun; Taipei City Psychiatric Center: MH Wu; Tzu Chi University: SY Chen; Shin Kong Wu Ho-Su Memorial Hospital: KE Chu; Huhsi Health Center, Penghu County: SC Ho, TG Lu; Provincial Penghu Hospital: WP Wu, TY Ou; Sanchi Health Center, Taipei County: CG Lin; Provincial Chutung Hospital: KC Shih; Provincial Potzu Hospital: WS Chung, C Li; Kaohsu Health Center, Pingtung County: CC Chen; Paihsa Health Center, Penghu County: WC How.

Editor’s quiz: GI snapshot ANSWER From the question on page 657 This patient has developed a ‘squashed stomach syndrome’. The CT shows encysted ascites within the lesser sac compressing the stomach such that it is reduced to a thin crescent. Lesser sac ascites is usually associated with metastatic ovarian cancer.1 In this case, the remainder of the CT examination revealed widespread abdominal and pelvic lymphadenopathy with deposits in the omental and pelvic fat. Bilateral heterogenous ovarian masses were subsequently biopsied under CT guidance and reported as poorly differentiated serous carcinoma. Squashed stomach syndrome is recognised in palliative care, often the result of hepatomegaly obstructing the gastric outlet.2 Obstruction of the upper gastrointestinal tract by lesser sac ascites, however, is rare, with only a few case reports.3 4 A study of 438 cases of ovarian cancer reported only three with squashed stomach syndrome due to ascitic compression.5 In these cases, image-guided drainage of the lesser sac relieved symptoms immediately.

694

Squashed stomach syndrome classically causes early satiety and gastro-oesophageal reﬂux, as in this case. Conservative management includes small, frequent meals and the use of metoclopramide.2 Competing interests None. Patient consent Obtained. Provenance and peer review Not commissioned; externally peer reviewed.

Gut 2011;60:694. doi:10.1136/gut.2009.190264a

REFERENCES 1. 2. 3. 4. 5.

Spencer JA. Ovarian cancer: what’s new, where next? Cancer Imaging 2003;4:19e21. Kuebler KK, Davis MP, Moore CD. Palliative practices. New York: Mosby, 2005. Katz LB, Frankel A, Cohen C, et al. Ovarian cancer complicated by gastric outlet obstruction. J Surg Oncol 1981;18:261e4. Krebs JB, Walsh J, Goplerud DR. Gastric outlet obstruction caused by ascitic fluid entrapment in the lesser sacda complication of advanced cancer: report of 2 cases. Gynecol Oncol 1982;14:105e11. Spencer JA, Crosse BA, Mannion RAJ, et al. Gastroduodenal obstruction from ovarian cancer: imaging features and clinical outcome. Clin Radiol 2000;55:264e72.

Gut May 2011 Vol 60 No 5

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology

Nanoliposomal ceramide prevents in vivo growth of hepatocellular carcinoma Hephzibah Rani S Tagaram,1 Nicole A DiVittore,2 Brian M Barth,2 James M Kaiser,2 Diego Avella,1 Eric T Kimchi,1 Yixing Jiang,3 Harriet C Isom,4 Mark Kester,2 Kevin F Staveley-O’Carroll1 1

Department of Surgery, Pennsylvania State University College of Medicine, Hershey, Pennsylvania, USA 2 Department of Pharmacology, Pennsylvania State University College of Medicine, Hershey, Pennsylvania, USA 3 Department of Medicine, Pennsylvania State University College of Medicine, Hershey, Pennsylvania, USA 4 Department of Microbiology and Immunology, Pennsylvania State University College of Medicine, Hershey, Pennsylvania, USA Correspondence to Mark Kester, Departments of Pharmacology and of Cellular and Molecular Physiology, 500 University Drive, R130, P O Box 850, Hershey, PA 17033-0858, USA; [email protected] Revised 8 October 2010 Accepted 8 November 2010 Published Online First 30 December 2010

ABSTRACT Background and objectives Hepatocellular carcinoma (HCC) affects an increasing number of people worldwide. The poor survival rate of patients with HCC is manifested by an aggressive and metastatic phenotype, as well as a poor response to common therapeutic strategies. The purpose of this study was to evaluate the efficacy of nanoliposomal C6-ceramide as an antineoplastic agent in an in vivo model of human HCC. Methods The growth-arresting and pro-apoptotic properties of nanoliposomal C6-ceramide were first evaluated in vitro in human SK-HEP-1 cells by assessing cellular viability, caspase 3/7 activity, annexin-V expression, DNA fragmentation, cell cycle distribution and AKT phosphorylation. SK-HEP-1 cells were then engrafted subcutaneously into athymic nude mice and nanoliposomal C6-ceramide was administered by tail vein injection. Tumour size was monitored over time, followed by excision of tumours to evaluate tumour vascularisation, proliferation, apoptosis and cellular signalling. Results Nanoliposomal C6-ceramide, but not ghost (no ceramide) nanoliposomes, induced apoptotic cell death of SK-HEP-1 cells in vitro, concomitant with an accumulation of cells in the G2 phase of the cell cycle and decreased phosphorylation of AKT. Systemic administration of nanoliposomal C6-ceramide to mice engrafted with SK-HEP-1 tumours reduced tumour vascularisation and proliferation, induced tumour cell apoptosis, decreased phosphorylation of AKT and ultimately blocked tumour growth. Conclusions These studies show that nanoliposomal ceramide is an efficacious antineoplastic agent for the treatment of in vitro and in vivo models of human HCC.

INTRODUCTION Hepatocellular carcinoma (HCC) is the ﬁfth most common cancer in the world with a 5-year survival rate of less than 5%.1e3 HCC is refractory to most chemotherapy, and radiation is often not possible due to liver toxicity.4e6 Treatment relies mainly on chemotherapy and/or surgical management via resection, transplantation or ablative techniques.7 8 Current treatment modalities including sorafenib offer limited survival beneﬁts; hence, there is a continuing need for innovative and alternative therapies for HCC.1 9 Ceramide, a bioactive sphingolipid, has been identiﬁed as a second messenger in cellular signalling cascades regulating differentiation, death and Gut 2011;60:695e701. doi:10.1136/gut.2010.216671

Significance of this study What is already known about this subject?

< The incidence of hepatocellular cellular carci-

noma continues to grow worldwide. antineoplastic chemotherapy, radiotherapy or surgical intervention is at best marginally effective for hepatocellular carcinoma.

< Conventional

What are the new findings?

< Insoluble and impermeable lipid-based antineo-

plastic chemotherapeutics can be administered systemically via nanotechnology. < Nanoliposomal ceramide can be used as an efficacious antineoplastic agent for the in vivo treatment of human hepatocellular carcinoma. < Systemic administration of nanoliposomal ceramide reduces tumour vascularisation, contributing to widespread tumour apoptosis and nearly complete abrogation of hepatocellular carcinoma growth.

How might it impact on clinical practice in the foreseeable future? < Based on these mechanistic studies, ceramide nanoliposomes are being developed for future clinical evaluation in patients with hepatocellular carcinoma.

senescence.10 11 It is generated naturally within cells via the hydrolysis of sphingomyelin or by de novo synthesis from palmitic acid.12 The levels of ceramide are increased within cells in response to several inducers of cellular stress including cytotoxic receptor systems, environmental stresses and various chemotherapeutics. Furthermore, pegylated nanoliposomal formulations of ceramide have demonstrated efﬁcacy in treating cancer via antiproliferative and pro-apoptotic properties both in vitro and in vivo.13e16 Mechanistically, nanoliposomal C6-ceramide has been shown to decrease endothelial markers of microvascularisation/angiogenesis at tumour sites15 and to decrease pro-survival (AKT) and pro-mitogenic (ERK) signalling pathways in tumours15 16 and endothelial cells.17 Importantly, nanoliposomal C6ceramide has been shown to be essentially free of toxic side effects normally associated with antineoplastic agents.18 Moreover, passively targeted pegylated nanoliposomal C6-ceramide has been 695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology shown to be selectively apoptotic for breast cancer but not normal mammary epithelial cells, as well as melanomas but not melanocytes.15 16 We investigated the anti-angiogenic and antineoplastic efﬁcacy of nanoliposomal C6-ceramide using in vitro and in vivo models of HCC and showed that the systemic administration of nanoliposomal C6-ceramide restricted angiogenesis and induced tumour cell apoptosis, ultimately preventing the growth of SK-HEP-1 tumour xenografts in athymic nude mice.

liposomes or nanoliposomal C6-ceramide for 24 h in media containing 1% FBS. Caspase 3/7 enzymatic activity was measured following treatment using an Apo-ONE Homogeneous Caspase 3/7 Assay according to the manufacturer’s instructions (Promega). Caspase 3/7 activity was determined by measuring the ﬂuorescence of the cleaved substrate using a microplate reader (excitation 498 nm; emission 521 nm).

Terminal deoxynucleotide transferase dUTP nick end labelling (TUNEL) assay

SK-HEP-1 human HCC cells were maintained at 378C and 5% CO2 in MEM supplemented with 10% fetal bovine serum (FBS), 1% sodium pyruvate, 1% non-essential amino acids, 1% sodium bicarbonate, 1% L-glutamine and 1% penicillin/streptomycin. For subculture, cells were subject to trypsin/EDTA detachment, centrifuged, resuspended in growth media and replated at appropriate cell density.

SK-HEP-1 cells were plated at 13104 cells per well in 8-well chamber slides and grown in 10% serum fortiﬁed media for 48 h prior to treatment. Cells were exposed to PBS, ghost nanoliposomes or nanoliposomal C6-ceramide for 24 h in media containing 1% FBS. Fragmented DNA of apoptotic cells was stained using an ApopTag Red In Situ Apoptosis Detection Kit according to the manufacturer ’s instructions (Chemicon, Temecula, California, USA) and visualised by ﬂuorescence microscopy using appropriate ﬁlters.

Liposome preparation

Annexin V assay

MATERIALS AND METHODS Cell culture

For nanoliposomal C6-ceramide, aliquots of 1,2-dioleoyl-sn-glycero-3-phosphocholine (DSPC), 1,2-dioleoylsn-glycero-3-phosphoethanolamine (DOPE), 1,2-distearoylsn-glycero-3-phosphoethanolamine-N-(methoxy(polyethylene gl ycol)-2000) (PEG2000-DSPE), C8-ceramide-1-succinyl[methoxy (polyethylene glycol)-750] and C6-ceramide suspended in chloroform were made in a 3.75:1.75:0.75:0.75:3 molar ratio. For ghost nanoliposomes, aliquots of DSPC, DOPE and PEG2000DSPE were made in a 5.66:2.87:1.47 molar ratio. In all experiments we evaluated SK-HEP-1 cells treated with phosphate buffer solution (PBS), a ghost nanoliposome formulation or a nanoliposomal C6-ceramide formulation. The ghost nanoliposomes contained the same amount of total lipid as the nanoliposomal C6-ceramide but no C6-ceramide. Lipids were dried to a ﬁlm under a stream of nitrogen, then hydrated by the addition of 0.9% NaCl to a ﬁnal lipid concentration of 25 mg/ml. Solutions were sealed, heated at 608C (60 min) and subjected to vortex mixing and sonicated until light no longer diffracted through the suspension. The lipid vesicle-containing solution was quickly extruded at 608C by passing the solution 10 times through 100 nm polycarbonate ﬁlters in an Avanti Mini-Extruder (Avanti Polar Lipids, Alabaster, Alabama, USA). Nanoliposome solutions were stored at 48C until use. Nanoliposomal size (80 nm) and a neutral charge were validated by a Zetasizer Nano ZS at 258C.

Cellular viability assay SK-HEP-1 cells were plated at 53103 cells per well in 96-well tissue culture plates and grown in 10% serum fortiﬁed media for 48 h prior to treatment. Cells were exposed to PBS, ghost nanoliposomes or nanoliposomal C6-ceramide for 24 h in media containing 1% FBS. Following treatment, cellular viability was assessed using a Cell Titre 96 AQueous Non-Radioactive Cell Proliferation Assay according to the manufacturer ’s instructions (Promega, Madison, Wisconsin, USA). Viability was determined by measuring absorbance at 490 nm using a microplate reader and normalising to the viability observed under control conditions.

SK-HEP-1 cells were plated at 13105 cells per well in 6-well tissue culture plates and grown in 10% serum fortiﬁed media for 48 h prior to treatment. Cells were exposed to PBS, ghost nanoliposomes or nanoliposomal C6-ceramide for 24 h in media containing 1% FBS. Cells were harvested and stained with PE-labelled annexin V (EMD Biosciences, San Diego, California, USA) and 7-aminoactinomycin D (Invitrogen, Carlsbad, California, USA) and analysed by ﬂow cytometry.

Cell cycle analysis SK-HEP-1 cells were grown in 10% serum fortiﬁed media until 50% conﬂuent. Cells were exposed to PBS, ghost nanoliposomes or nanoliposomal C6-ceramide for 24 h in regular growth media. Alternatively, cells were synchronised by 24 h serum starvation followed immediately by treatment or by full serum restoration for 10 h prior to treatment. Cells were dissociated from cultures plates and resuspended in PBS containing 50 mg/ml propidium iodide and 0.5 mg/ml RNase A. Flow cytometry and CellQuest software (BD Biosciences, San Jose, California, USA) was used to determine cell cycle distribution.

Western blotting Cell lysates were prepared in lysis buffer (0.1% NP40, 50 mM HEPES, 137 mM NaCl, 10 mM Na4P2O7, 50 mM NaF, 5 mM b-glycerolphosphate, 1 mM EGTA, 2 mM EDTA, 1% glycerol, 2 mM) containing protease inhibitor (EMD Biosciences) for 20 min on ice followed by centrifugation at 48C for 15 min to sediment particulate materials. Protein concentrations were measured using a Bio-Rad protein assay kit (Bio-Rad Laboratories, Hercules, California, USA). Proteins (30 mg) from whole cell extracts were separated on SDS-polyacrylamide gels and transferred onto nitrocellulose membranes. Membranes were blocked with 1% bovine serum albumin in TBS containing 0.05% Tween and incubated with phospho-AKT (Ser 473) and b-actin primary antibodies (Cell Signalling, Beverly, Massachusetts, USA) before visualisation with enhanced chemiluminescence detection (Thermo Scientiﬁc, Rockport, Illinois, USA).

In vivo tumour model Caspase 3/7 activity assay SK-HEP-1 cells were plated at 53103 cells per well in 96-well tissue culture plates and grown in 10% serum fortiﬁed media for 48 h prior to treatment. Cells were exposed to PBS, ghost nano696

Bilateral human HCC tumour xenografts were established in female athymic nude mice (Jackson Laboratories, Bar Harbour, Maine, USA) by subcutaneous injection of SK-HEP-1 cells over the rib cage. For each tumour, 53106 cells were resuspended in Gut 2011;60:695e701. doi:10.1136/gut.2010.216671

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology 200 ml cell culture media. Tumours were allowed to establish for 1 week prior to commencement of the treatment regime. Treatments occurred on alternate days via tail vein injection of sterile saline, ghost nanoliposomes or nanoliposomal C6-ceramide at 36 mg/kg body weight of the mice. Each treatment group consisted of ﬁve animals, each with bilateral tumours. Tumour volumes were quantiﬁed by measuring with callipers and multiplying tumour length, width and height. At the conclusion of the trial the animals were killed and the tumours were excised and processed for histological analysis by formalin ﬁxation, parafﬁn embedding and microtome sectioning. All animal procedures were approved by and carried out according to the standards and guidelines of the Pennsylvania State University College of Medicine Institutional Animal Care and Use Committee.

Tumour section fluorescence microscopy Tumour sections were deparafﬁnised in a series of hydrating solutions prior to staining. Markers of angiogenesis (speciﬁcally, endoglin/CD105, an endothelial cell surface glycoprotein linked to proliferation,19 and PECAM-1/CD31, an endothelial cell adhesion molecule regulating vessel formation19e21) were evaluated. Explant tissue sections were stained with CD31 and CD105 primary antibodies (eBiosciences, San Diego, California, USA). Tumour sections were also stained with a primary antibody detecting vascular endothelial growth factor (VEGF; Santa Cruz Biotechnology, Santa Cruz, California, USA) to further visualise tumour vascularisation. To quantify tumour cell apoptosis, explant tissue sections were ﬁrst treated with proteinase K (20 mg/ml, 15 min, 258C) and then an ApopTag Red In Situ Apoptosis Detection Kit was used according to the manufacturer ’s instructions (Chemicon). To evaluate tumour cell proliferation, the same sections were then stained with an antiproliferating cell nuclear antigen (PCNA) antibody conjugated to FITC (Santa Cruz Biotechnology). Tumour sections were alternatively stained with a phospho-AKT (Ser 473) primary antibody (Cell Signalling) to assess survival signalling. Appropriate ﬂuoroprobe-conjugated secondary antibodies were

used to detect CD31, CD105, VEGF or phospho-AKT. Cover slips were mounted using mounting medium containing DAPI nuclear stain. Stained tumour sections were visualised using a Leica TCS SP2 AOBS confocal microscope with the appropriate ﬁlters.

Statistical analysis One-way or two-way analysis of variance (ANOVA) were used to determine statistically signiﬁcant differences between treatments (p<0.05). At least three independent experiments were performed for each condition. Post hoc comparisons of speciﬁc treatments were performed using a Bonferroni test. All error bars represent SE from the mean. All statistical analyses were carried out using GraphPad Prism 4 software (La Jolla, California, USA).

RESULTS Nanoliposomal C6-ceramide induces SK-HEP-1 cell death in vitro We initially evaluated the in vitro efﬁcacy of nanoliposomal C6-ceramide for the treatment of human SK-HEP-1 cells, a model of metastatic HCC. Nanoliposomal C6-ceramide, PEGylated to improve retention and limit toxicity, has been shown to effectively treat cellular and animal models of breast cancer and melanoma.15 16 Nanoliposomal C6-ceramide induced a dose-dependent decrease in SK-HEP-1 cellular viability in vitro compared with PBS control and ghost nanoliposome treatment, as indicated by MTS viability assay (ﬁgure 1A). Furthermore, 10 mM nanoliposomal C6-ceramide, but not ghost nanoliposomes or PBS controls, stimulated a robust ﬁvefold increase in caspase 3/7 activity, indicative of apoptotic cell death (ﬁgure 1B). To further substantiate the apoptotic potential of nanoliposomal C6-ceramide in SK-HEP-1 cells, we assessed annexin V binding and TUNEL staining. Robust annexin V staining of SK-HEP-1 cells further indicated that 10 mM nanoliposomal C6-ceramide induced apoptotic cell death (ﬁgure 1C). Annexin V preferentially binds to phosphatidylserine, which is ﬂipped to the outer leaﬂet of the plasma membrane of cells undergoing apoptosis due to alterations in membrane

Figure 1 Nanoliposomal C6-ceramide decreases the viability of SK-HEP-1 cells in vitro by inducing apoptosis. Human SK-HEP-1 hepatocellular carcinoma cells were treated with 10 mM nanoliposomal C6-ceramide (C6), nanoliposomes without C6-ceramide (ghost) or phosphate buffer solution (control) in growth media supplemented with 1% fetal bovine serum. (A) Cellular viability was determined by MTS viability assay after 24 h treatment. (B) Caspase 3/7 activity was determined by fluorometric assay after 24 h treatment. (C) Annexin V binding (to externalised phosphatidylserine) was detected by flow cytometry after 24 h treatment. (D) DNA fragmentation was analysed by TUNEL staining after 24 h treatment. The TUNEL-positive control was treated with DNase. Data represent mean6SEM. *p<0.01. (A: n>6, two-way ANOVA; B: n¼2, one-way ANOVA; C: n¼3, one-way ANOVA). The images in (D) are representative of at least three independent experiments. Gut 2011;60:695e701. doi:10.1136/gut.2010.216671

697

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology asymmetry. Notably, the general addition of lipid in the form of ghost nanoliposomes exerted a modest degree of change in membrane asymmetry revealed by increased annexin V binding. TUNEL staining of SK-HEP-1 cells, which revealed DNA fragmentation, conﬁrmed that 10 mM nanoliposomal C6-ceramide, but not ghost nanoliposomes, induced apoptosis of SK-HEP-1 cells in vitro (ﬁgure 1D).

Nanoliposomal C6-ceramide induces SK-HEP-1 cell cycle arrest at either G1/S or G2/M checkpoints Cell cycle arrest is often a preliminary event that can lead to apoptosis. As ceramide has previously been reported to interfere with cell cycle progression in HCC,22 we evaluated the effect of nanoliposomal C6-ceramide on the cell cycle. We used ﬂow cytometry following DNA staining with propidium iodide to evaluate the cell cycle distribution of SK-HEP-1 cells following treatment with PBS, ghost nanoliposomes or nanoliposomal C6-ceramide (ﬁgure 2). With no serum manipulations, 10 mM nanoliposomal C6-ceramide induced a notable blockade of the cell cycle at the G2/M checkpoint. Interestingly, nanoliposomal C6ceramide treatment immediately following 24 h serum starvation to synchronise the cells also induced blockade of the cell cycle at the G1/S checkpoint. Additionally, SK-HEP-1 cell cycle progression following 24 h serum starvation was resumed by restoring regular growth media conditions for 10 h. Under this intermediate condition, nanoliposomal C6-ceramide induced G2/M blockade of only the cells which had entered the cell cycle, while the other cells were still blocked at G1/S. These results indicate that nanoliposomal C6-ceramide exerts an effect on the cell cycle

at a common point upstream of both the G1/S and G2/M checkpoints. Intriguingly, the cell survival kinase AKT is upstream of both checkpoints and has been previously shown to be regulated by ceramide.15 23

Nanoliposomal C6-ceramide decreases the phosphorylation of AKT AKT phosphorylation is often associated with resistance to apoptosis in cancer cells. The effect of nanoliposomal C6-ceramide on pro-survival AKT signalling was studied by western blot analysis. Following 24 h treatment with nanoliposomal C6-ceramide, phosphorylation of AKT was reduced in a dose-dependent manner (ﬁgure 3). Equal loading of lanes was conﬁrmed by b-actin staining. The results are consistent with previous studies showing that nanoliposomal C6-ceramide can target cancer cells by diminishing AKT signalling.15

Growth of SK-HEP-1 tumours in vivo is prevented by systemic administration of nanoliposomal C6-ceramide To evaluate the in vivo efﬁcacy of nanoliposomal C6-ceramide in a model of HCC, we established bilateral xenograft tumours in athymic nude mice. SK-HEP-1 cells were injected subcutaneously and tumours were allowed to establish for 1 week. Treatment commenced with a tail vein injection of nanoliposome formulations prepared in sterile isotonic saline. Mice received injections on alternate days with 36 mg/kg nanoliposomal C6-ceramide, an equivalent ghost nanoliposome concentration or isotonic saline. Tumour volume was monitored by measuring the length, width and height of the tumours

Figure 2 Nanoliposomal C6-ceramide promotes cell cycle arrest of SK-HEP-1 cells. SK-HEP-1 cells were synchronised by 24 h serum starvation (A). Treatments immediately followed cell synchronisation (B) or serum restoration for 10 h (C). Alternatively, treatments were conducted entirely under normal growth conditions with no serum starvation (D). Cells were treated with 10 mM nanoliposomal C6-ceramide (C6), nanoliposomes without C6-ceramide (ghost) or phosphate buffer solution (control) in growth media supplemented with 10% fetal bovine serum for an additional 24 h. Cells were collected, stained with propidium iodide and analysed by flow cytometry. DNA histograms are representative of three independent experiments. 698

Gut 2011;60:695e701. doi:10.1136/gut.2010.216671

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology

Figure 3 SK-HEP-1 cells treated with nanoliposomal C6-ceramide have diminished phosphorylation of AKT. SK-HEP-1 cells treated with 5, 10 and 20 mM nanoliposomal C6-ceramide, but not ghost nanoliposomes and control treatment, have significantly decreased phospho-AKT levels. b-actin and total AKT served as controls for protein loading. A representative blot of two experiments is shown. with callipers. Systemic administration of nanoliposomal C6-ceramide completely prevented the growth of human HCC xenografts in athymic nude mice whereas saline or ghost nanoliposomes did not (ﬁgure 4). Speciﬁcally, by the end of the trial (>7 weeks after tumour initiation), mice receiving saline or ghost nanoliposomes had tumour volumes over 3.5 times greater than those in mice treated with nanoliposomal C6-ceramide.

In vivo administration of nanoliposomal C6-ceramide restricts HCC tumour vascularisation and initiates apoptosis To conﬁrm the in vivo mechanisms by which nanoliposomal C6-ceramide prevents tumour growth, we assessed cellular proliferation and survival markers in excised tumours. SK-HEP-1 HCC tumours from mice treated systemically with either saline or ghost nanoliposomes had robust PCNA staining (ﬁgure 5A) and nearly no TUNEL-positive staining (ﬁgure 5A). However, tumours from mice treated systemically with nanoliposomal C6-ceramide had virtually no PCNA staining, yet dramatic

Figure 4 Nanoliposomal C6-ceramide prevents in vivo growth of SK-HEP-1 xenografted tumours in nude mice. Bilateral human SK-HEP-1 hepatocellular carcinoma xenografts were established in female athymic nude mice (five mice per treatment group). Systemic tail vein injections of sterile isotonic NaCl 36 mg/kg of ghost nanoliposomes (ghost) or 36 mg/kg nanoliposomal C6-ceramide (C6) were given on alternate days and tumour volume was determined by calliper measurement. The average tumour volume of the nanoliposomal C6-ceramide group was significantly different from the other treatment groups, as determined by two-way ANOVA. *p<0.05 vs control; #p<0.05 vs ghost. Gut 2011;60:695e701. doi:10.1136/gut.2010.216671

TUNEL-positive staining was observed. Consistent with increased apoptosis and decreased cellular proliferation, nanoliposomal C6-ceramide-treated animals but not ghost-treated or saline-treated animals had signiﬁcantly reduced phospho-AKT staining, a marker of pro-survival signalling (ﬁgure 5B). As ceramide has previously been shown to restrict angiogenesis24 which could also lead to tumour apoptosis, we evaluated tumour sections for CD31 and CD105, both markers of angiogenesis (ﬁgure 5C). Tumours from mice treated systemically with nanoliposomal C6-ceramide had dramatically less CD31 and CD105 staining than tumours from mice treated systemically with ghost nanoliposomes or saline control. Furthermore, signiﬁcant VEGF staining was observed in tumour sections from mice treated with ghost nanoliposomes or with saline, but not from those treated with nanoliposomal C6-ceramide (ﬁgure 5D). VEGF is a growth factor that promotes vascularisation. These ﬁndings suggest that the efﬁcacy of nanoliposomal C6-ceramide in vivo is also due to restriction of angiogenesis and therefore to diminished vascularisation of the tumours. Taken together, these results indicate that the efﬁcacy of nanoliposomal C6-ceramide is the result of a decrease in tumour vascularisation as well as a decrease in tumour cell proliferation and an increase in tumour cell apoptosis.

DISCUSSION The incidence of HCC continues to grow, possibly as a result of chronic liver disease which often precedes HCC. The long-term survival for patients with HCC is poor and the cancer often becomes metastatic before or during the course of treatment. Liver transplantation is currently a primary treatment option that can offer a cure if the cancer has not become metastatic, but leaves patients with a lifetime of taking immunosuppressive drugs to avoid transplant rejection.25 26 However, this surgical option remains limited owing to the scarcity of replacement organs as well as late diagnosis. No promising therapies exist for patients with advanced HCC. Although the multikinase inhibitor sorafenib has been found marginally to prolong survival in patients with advanced HCC, novel treatments are needed.27 Nanoliposomal C6-ceramide was designed as a therapeutic alternative to common chemotherapeutics. Liposomal formulation of short-chain ceramide analogues allows for the systemic delivery of this very hydrophobic cell-impermeable precipitating pro-apoptotic sphingolipid.16 28 29 Intriguingly, nanoliposomal C6-ceramide was shown to be toxic to cancer cells and not to normal cells, a paradigm arising from distinct ceramide metabolic regulation between these cells.13 16 Thus, the therapeutic utility of this cancer cell-selective antineoplastic agent depends on nanoscale delivery modalities. As a bioactive sphingolipid, ceramide has been linked to numerous cellular responses. Generally it is thought of as an inducer of cellular stress and apoptosis.11 30 Ceramide has also been shown speciﬁcally to regulate several enzymes which ultimately act as effectors. A prominent case is the activation of the atypical protein kinase C z which directly antagonises survival signalling through inhibitory phosphorylation of AKT.31 32 The present study implicates diminished phosphoAKT activity in the biological and therapeutic effects of exogenous nanoliposomal C6-ceramide. Given the role of AKT in cell survival, cell growth and cell cycle progression,23 it is not surprising that nanoliposomal C6-ceramide exerts such a robust antineoplastic effect including cell cycle blockade at both the G1/S and G2/M checkpoints. In fact, it is also plausible that diminished AKT activity induces the apoptotic phenotype in both HCC tumours and tumour vasculature, as has been shown 699

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Figure 5 Systemic administration of nanoliposomal C6-ceramide induces widespread apoptosis and decreased proliferation by decreasing pAKT activity in SK-HEP-1 tumours, concomitant with a reduction in markers of angiogenesis. Formalin-fixed paraffin-embedded SKHEP-1 tumour sections were deparaffinised and stained for markers of proliferation, apoptosis and angiogenesis. Tumour sections from mice treated with phosphate buffer solution, ghost nanoliposomes or nanoliposomal C6-ceramide were treated as follows: (A) stained with DAPI to reveal cellular nuclei, stained by the TUNEL method to reveal apoptotic cells, stained with PCNA to reveal proliferating cells; (B) stained for the presence of phospho-AKT; (C) stained for the presence of PECAM-1/CD31 and endoglin/CD105; and (D) stained for the presence of vascular endothelial growth factor. All were then visualised by confocal microscopy. Photomicrographic images are representative of three separate tumour sections per treatment.

in other systems.14e16 The documentation of decreased AKT activation in vivo ﬁrmly establishes the utility of targeting this pro-survival kinase with exogenous nanoliposomal ceramide. Many chemotherapeutics as well as radiation therapy have been shown to induce accumulations of ceramide.33 Speciﬁc to our study, we showed that nanoliposomal delivery of C6-ceramide in vivo to SK-HEP-1 HCC tumours blocked tumour vascularisation, which was visualised in tumour sections as a reduction in CD31 and CD105 staining as well as a reduction in VEGF staining. Tumour vascularisation or angiogenesis into and within tumours is necessary to establish nutrient support.34 It is not coincidental that studies have shown that nutrient deprivation can lead to cellular ceramide generation and cellular death.35 36 It is thus plausible that preventing HCC tumour vascularisation with systemic nanoliposomal C6-ceramide treatment could induce widespread apoptosis within the tumours. In fact, short-chain ceramide treatment has been shown to generate more physiological long-chain ceramide species within tumours.37 The efﬁcacy of our nanoliposomal C6-ceramide in the treatment of diverse cancers lends signiﬁcant therapeutic promise as it translates from the bench to the bedside. Rather than stimulating ceramide endogenously, we have administered ceramide exogneously, overcoming solubility issues and maintaining a non-toxic proﬁle to normal cells.18 Ceramide nanoliposomes are speciﬁcally formulated so that a maximal amount of ceramide (30 molar per cent) is intercalated within the bilayer. Furthermore, these nanoliposomes are PEGylated to improve systemic retention, improve enhanced permeability retention in tumours and minimise the recognition and clearance by the host immune system. Together, these characteristics distinguish ceramide nanoliposomes and promote their potential as an efﬁcacious antineoplastic agent. This study has further shown that nanoliposomal C6-ceramide is efﬁcacious in treating in vitro 700

and in vivo human models of HCC. In addition to direct apoptotic effects of ceramide towards tumour cells, ceramide nanoliposomes further reduced tumour vascularisation, contributing to widespread tumour cell apoptosis and ultimately to an abrogation of in vivo HCC tumour growth. This study provides direct evidence that nanoliposomal C6-ceramide is efﬁcacious in the treatment of HCC. Acknowledgements The authors thank Nate Scheaffer and Dr David R Stanford of the flow cytometry for their assistance during the course of this study. Funding This work was funded in part by a Penn State Deans Feasibility Grant to HRT, by State of Pennsylvania non-formulary Tobacco Funds to MK, NIH R01 HL076789 to MK and K08 417-67HY to KFS. Competing interests The Penn State Research Foundation has licensed ceramide nanoliposomes and other nonliposomal nanotechnology to Keystone Nano Inc (State College, PA). MK is the Chief Medical Officer of Keystone Nano Inc. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4. 5.

6. 7.

Shariff MI, Cox IJ, Gomaa AI, et al. Hepatocellular carcinoma: current trends in worldwide epidemiology, risk factors, diagnosis and therapeutics. Expert Rev Gastroenterol Hepatol 2009;3:353e67. Cabrera R, Nelson DR. Review article: the management of hepatocellular carcinoma. Aliment Pharmacol Ther 2010;31:461e76. Singal AG, Marrero JA. Recent advances in the treatment of hepatocellular carcinoma. Curr Opin Gastroenterol 2010;26:189e95. Lee JO, Kim DY, Lim JH, et al. Palliative chemotherapy for patients with recurrent hepatocellular carcinoma after liver transplantation. J Gastroenterol Hepatol 2009;24:800e5. Uhm JE, Park JO, Lee J, et al. A phase II study of oxaliplatin in combination with doxorubicin as first-line systemic chemotherapy in patients with inoperable hepatocellular carcinoma. Cancer Chemother Pharmacol 2009;63:929e35. Seong J. Challenge and hope in radiotherapy of hepatocellular carcinoma. Yonsei Med J 2009;50:601e12. Sotiropoulos GC, Malago M, Molmenti EP, et al. Impact of a living donor liver transplant program for hepatocellular carcinoma to the running local deceased donor liver transplant program. Dig Dis Sci 2010;55:872e3.

Gut 2011;60:695e701. doi:10.1136/gut.2010.216671

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23.

Chua TC, Liauw W, Saxena A, et al. Systematic review of neoadjuvant transarterial chemoembolization for resectable hepatocellular carcinoma. Liver Int 2009;30:166e74. Spangenberg HC, Thimme R, Blum HE. Evolving therapies in the treatment of hepatocellular carcinoma. Biologics 2008;2:453e62. Kolesnick RN, Kronke M. Regulation of ceramide production and apoptosis. Annu Rev Physiol 1998;60:643e65. Hannun YA, Obeid LM. Ceramide: an intracellular signal for apoptosis. Trends Biochem Sci 1995;20:73e7. Dobrowsky RT, Jenkins GM, Hannun YA. Neurotrophins induce sphingomyelin hydrolysis. Modulation by co-expression of p75NTR with Trk receptors. J Biol Chem 1995;270:22135e42. Shabbits JA, Mayer LD. Intracellular delivery of ceramide lipids via liposomes enhances apoptosis in vitro. Biochim Biophys Acta 2003;1612:98e106. Stover T, Kester M. Liposomal delivery enhances short-chain ceramide-induced apoptosis of breast cancer cells. J Pharmacol Exp Ther 2003;307:468e75. Stover TC, Sharma A, Robertson GP, et al. Systemic delivery of liposomal short-chain ceramide limits solid tumor growth in murine models of breast adenocarcinoma. Clin Cancer Res 2005;11:3465e74. Tran MA, Smith CD, Kester M, et al. Combining nanoliposomal ceramide with sorafenib synergistically inhibits melanoma and breast cancer cell survival to decrease tumor development. Clin Cancer Res 2008;14:3571e81. Niwa S, Graf C, Bornancin F. Ceramide kinase deficiency impairs microendothelial cell angiogenesis in vitro. Microvasc Res 2009;77:389e93. Zolnik BS, Stern ST, Kaiser JM, et al. Rapid distribution of liposomal short-chain ceramide in vitro and in vivo. Drug Metab Dispos 2008;36:1709e15. Xiong YQ, Sun HC, Zhang W, et al. Human hepatocellular carcinoma tumor-derived endothelial cells manifest increased angiogenesis capability and drug resistance compared with normal endothelial cells. Clin Cancer Res 2009;15:4838e46. Zhao XL, Du J, Zhang SW, et al. [A study on vasculogenic mimicry in hepatocellular carcinoma]. Chung Hua.Kan Tsang Ping Tsa Chih 2006;14:41e4. Sun B, Zhang S, Zhang D, et al. Vasculogenic mimicry is associated with high tumor grade, invasion and metastasis, and short survival in patients with hepatocellular carcinoma. Oncol Rep 2006;16:693e8. Wang J, Lv XW, Shi JP, et al. Mechanisms involved in ceramide-induced cell cycle arrest in human hepatocarcinoma cells. World J Gastroenterol 2007;13:1129e34. Liang J.Slingerland JM. Multiple roles of the PI3K/PKB (Akt) pathway in cell cycle progression. Cell Cycle 2003;2:339e45.

Gut 2011;60:695e701. doi:10.1136/gut.2010.216671

24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

Li H, Junk P, Huwiler A, et al. Dual effect of ceramide on human endothelial cells: induction of oxidative stress and transcriptional upregulation of endothelial nitric oxide synthase. Circulation 2002;106:2250e6. Kornberg A, Kupper B, Tannapfel A, et al. Long-term survival after recurrent hepatocellular carcinoma in liver transplant patients: clinical patterns and outcome variables. Eur J Surg Oncol 2010;36:275e80. Lao OB, Weissman J, Perkins JD. Pre-transplant therapy for hepatocellular carcinoma is associated with a lower recurrence after liver transplantation. Clin Transplant 2009;23:874e81. Kerr SH, Kerr DJ. Novel treatments for hepatocellular cancer. Cancer Lett 2009;286:114e20. Saddoughi SA, Song P, Ogretmen B. Roles of bioactive sphingolipids in cancer biology and therapeutics. Subcell Biochem 2008;49:413e40. Heakal Y, Kester M. Nanoliposomal short-chain ceramide inhibits agonist-dependent translocation of neurotensin receptor 1 to structured membrane microdomains in breast cancer cells. Mol Cancer Res 2009;7:724e34. Basu S, Kolesnick R. Stress signals for apoptosis: ceramide and c-Jun kinase. Oncogene 1998;17:3277e85. Bourbon NA, Sandirasegarane L, Kester M. Ceramide-induced inhibition of Akt is mediated through protein kinase C zeta: implications for growth arrest. J Biol Chem 2002;277:3286e92. Fox TE, Houck KL, O’Neill SM, et al. Ceramide recruits and activates protein kinase C zeta (PKC zeta) within structured membrane microdomains. J Biol Chem 2007;282:12450e7. Ogretmen B, Hannun YA. Biologically active sphingolipids in cancer pathogenesis and treatment. Nat Rev Cancer 2004;4:604e16. Mesri M, Morales-Ruiz M, Ackermann EJ, et al. Suppression of vascular endothelial growth factor-mediated endothelial cell protection by survivin targeting. Am J Pathol 2001;158:1757e65. Jayadev S, Liu B, Bielawska AE, et al. Role for ceramide in cell cycle arrest. J Biol Chem 1995;270:2047e52. Colombaioni L, Frago LM, Varela-Nieto I, et al. Serum deprivation increases ceramide levels and induces apoptosis in undifferentiated HN9.10e cells. Neurochem Int 2002;40:327e36. Ogretmen B, Pettus BJ, Rossi MJ, et al. Biochemical mechanisms of the generation of endogenous long chain ceramide in response to exogenous short chain ceramide in the A549 human lung adenocarcinoma cell line. Role for endogenous ceramide in mediating the action of exogenous ceramide. J Biol Chem 2002;277:12960e9.

701

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology

Working Party proposal for a revised classification system of renal dysfunction in patients with cirrhosis Florence Wong,1 Mitra K Nadim,2 John A Kellum,3 Francesco Salerno,4 Rinaldo Bellomo,5 Alexander Gerbes,6 Paolo Angeli,7 Richard Moreau,8 Andrew Davenport,9 Rajiv Jalan,10 Claudio Ronco,11 Yuri Genyk,12 Vicente Arroyo13 1

Department of Medicine, University of Toronto, Toronto, Canada 2 Division of Nephrology, University of Southern California, Los Angeles, California, USA 3 Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, USA 4 Policlinico IRCCS San Donato and Dipartimento di Scienze Medico-Chirurgiche, Universita` di Milano, Milano, Italy 5 Department of Intensive Care, Austin Health, Melbourne, Victoria, Australia 6 Liver Center Munich, Klinikum Mu¨nchen-Grosshadern, Ludwig-Maximilians-Universita¨t Mu¨nchen, Mu¨nchen, Germany 7 Department of Clinical and Experimental Medicine, University of Padova, Padova, Italy 8 Centre de Recherche Biome´dicale Bichat-Beaujon, Service d’He´patologie, Hoˆpital Beaujon, Clichy, France 9 Centre for Nephrology, University College London Medical School, London, UK 10 Institute of Hepatology, University College London Medical School, London, UK 11 Division of Nephrology, Dialysis and Transplantation, San Bortolo Hospital, Vicenza, Italy 12 Division of Multiorgan Abdominal Transplantation, University of Southern California, Los Angeles, California, USA 13 Liver Unit, Institute of Digestive and Metabolic Diseases, Hospital Clinic, University of Barcelona, Spain Correspondence to Florence Wong, 9N/983, Department of Medicine, Toronto General Hospital, University of Toronto, 200 Elizabeth Street, Toronto, Ontario M5G2C4, Canada; [email protected] Revised 22 January 2011 Accepted 24 January 2011 Published Online First 15 February 2011

702

ABSTRACT Objectives To propose an improvement on the current classification of renal dysfunction in cirrhosis. Clinicians caring for patients with cirrhosis recognize that the development of renal dysfunction is associated with significant morbidity and mortality. While most cases of renal dysfunction in cirrhosis are functional in nature, developed as a result of changes in haemodynamics, cardiac function, and renal auto-regulation, there is an increasing number of patients with cirrhosis and structural changes in their kidney as a cause of renal dysfunction. Therefore, there is a need for a newer classification to include both functional and structural renal diseases. Design A working party consisting of specialists from multiple disciplines conducted literature search and developed summary statements, incorporating the renal dysfunction classification used in nephrology. These were discussed and revised to produce this proposal. Setting Multi-disciplinary international meeting. Patients None. Interventions Literature search using keywords of cirrhosis, renal dysfunction, acute kidney injury (AKI), chronic kidney injury (CKD), and hepatorenal syndrome. Results Acute kidney injury will include all causes of acute deterioration of renal function as indicated by an increase in serum creatinine of >50% from baseline, or a rise in serum creatinine of $26.4mol/L ($0.3mg/dL) in <48hours. Chronic renal disease will be defined as an estimated glomerular filtration rate (GFR) of <60ml/min calculated using the Modification of Diet in Renal Disease 6 (MDRD6) formula, recognising that the MDRD6 formula is not perfect for the cirrhotic patients and this may change as improved means of estimating GFR becomes available. Acute on chronic kidney disease will be defined as AKI superimposed on existing chronic renal disease using the above definitions for AKI and CKD. Conclusions Accepting this new classification will allow studies into the epidemiology, incidence, prevalence, natural history and the development of new treatments for these subtypes of renal dysfunction in cirrhosis.

Acute kidney injury (AKI) is common in patients with cirrhosis and ascites, occurring in up to 19% of cirrhotic patients admitted to hospital. In addition, chronic kidney disease (CKD) occurs in approximately 1% of all patients with cirrhosis.1 The combination of liver disease and renal dysfunction can occur as a result of systemic conditions that affect both the liver and the kidney simultaneously. However, renal dysfunction complicating primary disorders of the liver are much more common. These may include structural

Significance of this study What is already known about this subject?

< Hepatorenal syndrome is a severe complication

of advanced cirrhosis with a poor prognosis if left untreated. < The diagnosis of hepatorenal syndrome requires the patient fulfilling a set of diagnostic criteria. < Once a diagnosis of hepatorenal syndrome is made, treatments are available and these are effective in up to 40% of patients.

What are the new findings? < A proposal to broaden the diagnosis of renal

dysfunction in cirrhosis to include cases of acute and chronic renal failure not meeting the diagnostic criteria of hepatorenal syndrome types 1 and 2, respectively. < Acute kidney injury will include all causes of acute deterioration of renal function as indicated by an increase in serum creatinine of >50% from baseline or a rise in serum creatinine of $26.4 mmol/l ($0.3 mg/dl) in <48 h. < Chronic renal disease will be defined as an estimated glomerular filtration rate of <60 ml/ min for more than 3 months calculated using the Modification of Diet in Renal Disease 6 formula. < Acute on chronic kidney disease will be defined as an acute kidney injury superimposed on existing chronic renal disease using the above definitions for acute kidney injury and chronic kidney disease.

How might it impact on clinical practice in the foreseeable future? < The recognition of cases of renal dysfunction outside the traditional definition of hepatorenal syndrome will allow patients with lesser degrees of renal dysfunction to receive treatment. < The acceptance of these broadened definitions of renal dysfunction in cirrhosis will help to design studies to assess the pathophysiology, and thence to devise treatment strategies for these patients. < A better classification system may also secure more correct diagnoses leading to earlier and better treatment. < This potentially could have a positive impact on patient outcome, as patients will be treated earlier in the natural history of renal dysfunction.

Gut 2011;60:702e709. doi:10.1136/gut.2010.236133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology renal diseases such as IgA nephropathy, membranous nephropathy and cryoglobulinaemia, or renal dysfunction without signiﬁcant histopathological changes such as hepatorenal syndrome (HRS). These episodes of renal dysfunction may occur acutely and are associated with signiﬁcant morbidity and mortality. With improved understanding of renal complications in cirrhosis and the advent of treatment options, there is now a greater need to diagnose renal dysfunction in cirrhosis accurately. The Acute Dialysis Quality Initiative (ADQI) is an ongoing process that seeks to produce evidence-based recommendations for the prevention and management of AKI.2 As AKI has not been formally deﬁned in patients with cirrhosis, members of the ADQI and the International Ascites Club (IAC) formed a Working Group in March 2010 to discuss the deﬁnition of renal dysfunction (both acute and chronic) in patients with cirrhosis. Members of the Working Group included specialists who are experts in the pathophysiology and management of renal dysfunction in cirrhosis and were selected from the membership of the ADQI and IAC. They conducted a literature search and developed summary statements which were discussed and revised at the meeting. The participants of the joint ADQIeIAC meeting are shown in appendix 1. The ﬁnal summary statements and directions for future research are the basis for this paper.

HISTORICAL PERSPECTIVE The clinical entity we now know as HRS was originally described by Flint in 1963.3 In 1959, Papper et al reported intense renal vasoconstriction in an otherwise normal kidney in such patients, paving the way for the understanding of the pathogenesis of HRS.4 Epstein et al later conﬁrmed renal vasoconstriction using renal angiography in a patient with cirrhosis dying from renal failure and demonstrated post-mortem ﬁlling of all renal vessels to the periphery of the cortex, thus establishing the ‘functional nature’ of HRS.5 Rodes et al next identiﬁed three different outcome patterns in cirrhotic patients with renal dysfunction6: (1) a rapidly progressive course with a history of a complication closely related to the onset of renal failure (this group was later classiﬁed as type 1 HRS); (2) patients with stable renal dysfunction during hospitalisation but no obvious cause for renal failure (type 2 HRS); and (3) patients with an initial similar course as those in group 2 until some complication occurred that hastened the course of renal failure. The outcome was worst for patients in the ﬁrst group and best for patients in the second group.

nous vasoactive systems.9 Clinically, HRS was divided into two types: type 1 or acute HRS was characterised by a rapidly progressive reduction of renal function as deﬁned by a doubling of the initial serum creatinine to >220 mmol/l (2.5 mg/dl) or a 50% reduction in the initial 24 h creatinine clearance to <20 ml/min in <2 weeks; type 2 or chronic HRS was deﬁned as moderate renal failure that progressed gradually over weeks to months with a serum creatinine of 133e220 mmol/l (1.5e2.5 mg/dl). The IAC updated the deﬁnition and diagnostic criteria for HRS in 2005 (box 1).10 This came about because of an improved understanding of the pathophysiology of HRS, the recognition that it frequently follows bacterial infections (especially spontaneous bacterial peritonitis), the development of effective treatments and improved survival for patients with HRS, especially type 1. HRS is therefore no longer necessarily a fatal condition without liver transplantation.

PATHOPHYSIOLOGY OF HEPATORENAL SYNDROME The following is a summary of the current understanding of the pathophysiology of HRS (ﬁgure 1).

Portal hypertension as the initiator of haemodynamic changes The development of cirrhosis is associated with distortion, compression and even obliteration of the liver vasculature. In addition, there is decreased intrahepatic production of vasodilators and activated hypercontractile stellate cells.11 This overall increased resistance to portal inﬂowdor portal hypertension12dwill increase the shear stress on the splanchnic vessel walls leading to increased production of various vasodilators such as nitric oxide, causing splanchnic vasodilation.13 Several other factors including increased bacterial translocation, increased mesenteric angiogenesis and hyporesponsiveness of the splanchnic vessels to vasoconstrictors also contribute to the splanchnic vasodilation.14 The end result is a pooling of blood in the splanchnic vascular bed, akin to a splanchnic steal syndrome.15 The shunting of blood and excess vasodilators from the splanchnic to the systemic circulation following the opening of portal-systemic shunts related to increased portal pressure also leads to systemic arterial vasodilation.16 The combined effect causes a relative inadequacy of the systemic circulation, the so-called ‘reduction in the effective arterial blood volume’, thereby triggering a hyperdynamic circulation in these patients.17 18 Independent of these haemodynamic changes, portal hypertension per se can lead to renal vasoconstriction via increased

DEFINITION OF HEPATORENAL SYNDROME In 1979, a group of international investigators deﬁned HRS as a progressive form of renal dysfunction that occurred in cirrhosis and other severe parenchymal liver diseases,7 with features of prerenal renal failure (low urine sodium concentration and hyperosmolar urine) but without any improvement following volume expansion. However, they recognised that some cases do progress to acute tubular necrosis. Despite setting guidelines, there continued to be confusion over what truly constituted HRS. This led to an editorial in the Lancet8 suggesting the term ‘hepatic nephropathy’ to distinguish functional renal failure from any combination of renal failure occurring with liver failure, such as paracetamol overdose causing combined liver and renal failure. In 1996, the IAC deﬁned HRS as a syndrome that occurs in patients with cirrhosis, portal hypertension and advanced liver failure, characterised by impaired renal function with marked abnormalities in the arterial circulation and activity of endogeGut 2011;60:702e709. doi:10.1136/gut.2010.236133

Box 1 International Ascites Club (IAC) proposed diagnostic criteria for hepatorenal syndrome10 < Cirrhosis with ascites < Serum creatinine >133 mmol/l (1.5 mg/dl) < No improvement in serum creatinine (decrease to a level of

#133 mmol/l or 1.5 mg/dl) after at least 2 days of diuretic withdrawal and volume expansion with albumin. The recommended dose of albumin is 1 g/kg body weight/day up to a maximum of 100 g/day < Absence of shock < No current or recent treatment with nephrotoxic drugs < Absence of parenchymal kidney disease as indicated by proteinuria >500 mg/day, microhaematuria (>50 red blood cells/high power field) and/or abnormal renal ultrasonography

703

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Figure 1 Pathophysiology of hepatorenal syndrome: acute precipitating event. HRS, hepatorenal syndrome.

sympathetic nervous activity. For example, the elimination of portal hypertension with the insertion of a transjugular intrahepatic portosystemic shunt (TIPS) is able to improve renal blood ﬂow19 associated with a reduction in sympathetic nervous activity.20 The infusion of glutamine which increases hepatic sinusoidal pressure, mimicking portal hypertension, reduces the glomerular ﬁltration rate (GFR).21 Finally, lumbar sympathetic blockade in patients with HRS increases renal blood ﬂow, suggesting that the renal sympathetic activity is implicated in the efferent arm of this hepatorenal reﬂex.22

Excess renal vasoconstriction A reduced effective arterial blood volume results in the compensatory activation of various vasoconstrictor systems. In response, renal blood ﬂow decreases with consequent reduction in GFR. Normally, the kidneys maintain blood ﬂow by increasing production of renal vasodilators such as prostaglan704

dins and kallikrein. However, in patients with cirrhosis there is an overall reduction in renal vasodilator production,15 23 thereby favouring renal vasoconstriction.24 This renal hypoperfusion further increases the production of various intrarenal vasoconstrictors including angiotensin II and endothelin, causing further deterioration of renal haemodynamics and renal function, occasionally with glomerular ischaemia and mesangial constriction.25

Abnormal renal autoregulation Renal autoregulation is the process whereby regulatory mechanisms ensure that the kidneys receive a relatively constant blood supply regardless of ﬂuctuations in blood pressure. Below a critical threshold of 65 mm Hg, renal blood ﬂow decreases in proportion to renal perfusion pressure which, in turn, is dependent on mean arterial pressure. In cirrhosis, there is a progressive rightward shift of the renal autoregulation curve as liver disease progressesdthat is, for every given renal perfusion Gut 2011;60:702e709. doi:10.1136/gut.2010.236133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology pressure, there is a gradual reduction of renal blood ﬂow as liver disease advances.26 The patient with cirrhosis is therefore poised to develop renal failure simply because of the presence of advanced cirrhosis.

Abnormal cardiocirculatory function The high cardiac output state of the hyperdynamic circulation in decompensated cirrhosis means that there is limited cardiac reserve in these patients, and further reductions in systemic vascular resistance cannot be met with further increases in cardiac output. Failure to maintain blood pressure further compromises renal perfusion. In cirrhotic patients with ascites and spontaneous bacterial peritonitis, and therefore further arterial vasodilation as a result of the infection, those who went on to develop HRS at infection resolution had signiﬁcantly lower cardiac output compared with baseline and also compared with those who did not develop HRS. A relative inability to increase cardiac output during stress, a condition known as cirrhotic cardiomyopathy,27 28 may therefore be a risk factor for the development of HRS.29 Indeed, a relative low cardiac output and high plasma renin activity were signiﬁcant predictors for the development of HRS in cirrhosis with ascites.30 The fact that blockage of a TIPS shunt with an angioplasty balloon instantly reduces renal blood ﬂow, which reverses upon deﬂation of the balloon, conﬁrms that a reduction or increase in venous return and hence cardiac output has a direct bearing on renal haemodynamics.31 Recently, the relationship between cardiac systolic dysfunction and the risk of developing renal dysfunction in cirrhosis was also conﬁrmed, as well as the negative impact of cardiac dysfunction on patient survival.32 All the above factors contribute to the gradual deterioration in renal function as cirrhosis advances. Any event that causes an abrupt deterioration in haemodynamics can lead to a rapid decline in renal function, precipitating type 1 HRS (ﬁgure 1).

CURRENT DIAGNOSTIC CRITERIA FOR HEPATORENAL SYNDROME: ADVANTAGES AND DISADVANTAGES The most recent diagnostic criteria for HRS clearly delineated which patients should be regarded as having HRS and therefore receive speciﬁc treatment. However, the rigid cut-off value of a serum creatinine level of 133 mmol/l (1.5 mg/dl) may limit treatment to patients with the most severe degree of renal dysfunction. The changes that predispose to the development of HRS are not an ‘all-or-none’ phenomenon, but rather evolve progressively with the natural history of cirrhosis (ﬁgure 2). It is unclear whether patients who have milder degrees of renal dysfunction will also experience adverse outcomes. If so, they should also be offered treatment early rather than waiting until the diagnostic criteria of HRS are reached. Additionally, serum

Figure 2 Natural history of cirrhosis: acute precipitating event. AKI, acute kidney injury; GFR, glomerular filtration rate. Gut 2011;60:702e709. doi:10.1136/gut.2010.236133

creatinine is notoriously inaccurate in the diagnosis of renal dysfunction in cirrhosis.33 Although serum creatinine reﬂects renal function in patients with compensated cirrhosis fairly accurately, patients with decompensated cirrhosis often have low serum creatinine levels relative to their GFR owing to reduced production of creatinine from creatine in the liver and signiﬁcant muscle wasting.34 Thus, serum creatinine in patients with decompensated cirrhosis can still be within the normal range despite signiﬁcant renal dysfunction.35 The use of creatinine clearance in cirrhosis to assess renal function is also unreliable because of the falsely low serum creatinine in these patients coupled with a relatively increased renal tubular creatinine secretion compared with ﬁltered creatinine. Furthermore, it requires a 24-h urine collection which is often incomplete. Formulae such as the CockcrofteGault and Modiﬁcation of Diet in Renal Disease (MDRD)dwhich are based on serum creatinine concentrationsdwill also overestimate the GFR in cirrhosis.36 37 Clearance techniques using exogenous markers such as inulin or iothalamate provide a more accurate measurement of GFR but are labour-intensive and expensive.38 The use of a one-sample 51Cr-EDTA clearance technique is much simpler. However, this method tends to overestimate true renal function in patients with both volume overload and ascites due to redistribution of tracer into the ascitic and interstitial ﬂuid. These problems in the estimation of GFR are compounded by correcting for body surface area.39 Other biological markers such as cystatin C40 and neutrophil gelatinase associated lipocalin (NGAL),41 although promising, have not been validated in patients with advanced liver disease. Therefore, until better measurements of GFR can be found and validated, serum creatinine measurement remains the most widely used method for estimating renal function in clinical practice in patients with cirrhosis.42 Recognising the inadequacy of serum creatinine as an index of renal function in cirrhosis, patients with milder degrees of renal dysfunction may not be diagnosed until advanced renal failure sets in. The ADQIeIAC Working Group therefore proposes the following deﬁnitions for the diagnosis of renal dysfunction in cirrhosis in order to help identify patients with milder renal dysfunction for possible treatment. Since no studies have been performed in cirrhosis using these proposed deﬁnitions, they can best be regarded as expert opinions or level D evidence, but they represent an important ﬁrst step in the process of standardising nomenclature and deﬁnitions in patients with cirrhosis and renal dysfunction. It is planned that this empirical proposed classiﬁcation will be validated in prospective trials.

DEFINITION OF ACUTE KIDNEY INJURY IN CIRRHOSIS In 2004 the ADQI Working Group developed a consensus deﬁnition and classiﬁcation for AKI known as the RIFLE criteria (R: renal risk, I: injury, F: failure, L: loss of kidney function, E: endstage renal disease) which stratiﬁed acute renal dysfunction into grades of increasing severity based on changes in serum creatinine and/or urine output (ﬁgure 3).43 To date, the RIFLE criteria have been validated in over 500 000 patients with AKI44 45 and have been shown to predict clinical outcomes with a progressive increase in mortality with worsening RIFLE class.46 The Acute Kidney Injury Network (AKIN), an independent collaborative network consisting of experts from ADQI and several nephrology and intensive care medicine societies, broadened the deﬁnition of AKI to include an absolute increase in serum creatinine of $26 mmol/l ($0.3 mg/dl) when documented to occur within 48 h,47 since smaller increases in serum creatinine than those considered in the RIFLE classiﬁcation have been 705

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology

Figure 3 The RIFLE (R: renal risk, I: injury, F: failure, L: loss of kidney function, E: end-stage renal disease) diagnostic criteria.43 ARF, acute renal failure; GFR, glomerular filtration rate; UO, urine output. shown to be associated with an adverse outcome.48 Once established, a staging system then deﬁnes the severity of the AKI (table 1). The spectrum of kidney disease in cirrhosis includes acute and chronic conditions. Nephrologists distinguish acute and chronic renal disease by an artiﬁcial timeline of 3 months. Using the RIFLE/AKIN criteria for AKI, only a few patients with cirrhosis and acute kidney dysfunction will meet the criteria for type 1 HRS and therefore the remainder will have to be regarded as having AKI, be it structural or functional. Similarly, some patients with cirrhosis will have CKD such as diabetic nephropathy or mild renal dysfunction not reaching a serum creatinine of 133 mmol/l (1.5 mg/dl), and therefore not meeting the criteria for a diagnosis of type 2 HRS. HRS therefore only describes a portion of cirrhotic patients with renal dysfunction. The ADQIeIAC proposed that the term ‘hepatorenal disorders’ (HRD) be used to describe all concurrent kidney dysfunction in patients with advanced liver diseasedwhether functional or structural in naturedwhich fulﬁls the diagnostic criteria of AKI or CKD or HRS (ﬁgure 4). Such a deﬁnition is not meant to replace the current deﬁnition of HRS, but rather to be inclusive of all patients with renal dysfunction so that a proper classiﬁcation of renal dysfunction and appropriate studies can be conducted to deﬁne their prognosis and to devise treatment options. Using the creatinine criteria for AKI in patients with cirrhosis will certainly identify many patients with acute renal dysfunction and normal serum creatinine but low GFR. The urine output criteria for AKI may not be applicable in cirrhosis since patients with refractory ascites may maintain a urine output of <0.5 ml/kg/h even in the absence of AKI. The ﬁnal consensus proposal of the Working Party was to accept the deﬁnition of AKI in cirrhosis as an increase in serum creatinine of >50% from baseline or a rise in serum creatinine of $26.4 mmol/l ($0.3 mg/dl) in <48 h, irrespective of whether the cause of the acute deterioration in renal function is related to a functional or structural disorder (table 2). Type 1 HRS can be regarded as

a speciﬁc form of AKI. It was further agreed that these empirical new diagnostic criteria of AKI for cirrhosis will be validated to determine whether these smaller increases in serum creatinine are associated with poor outcomes. Two studies involving critically ill patients with cirrhosis admitted into an intensive care unit already showed that the RIFLE criteria for AKI was a good predictor of hospital survival.49 50 Once conﬁrmed, the serum creatinine threshold for the diagnosis of type 1 HRS may need to be revised to a lower target value. This has the potential to allow patients with a smaller rise in creatinine to beneﬁt from treatments currently reserved for patients with classical HRS. This new classiﬁcation will also allow studies of the epidemiology, incidence, prevalence and natural history of various subtypes of AKI in cirrhosis, thereby allowing the development of potential preventive and treatment strategies.

DEFINITION OF CHRONIC KIDNEY DISEASE IN CIRRHOSIS Patients with chronic renal impairment related to cirrhosis may not ﬁt the deﬁnition and staging of CKD (table 3) as set out by the practice guidelines from the Kidney Disease Outcomes Quality Initiatives (K/DOQI) Workgroup,51 since it requires a GFR of <60 ml/min/1.73 m2 for >3 months, irrespective of the presence or absence of structural kidney damage. As mentioned above, estimation of GFR in cirrhosis using various formulae is problematic and actual measurement of GFR using iothalamate or inulin clearance techniques are cumbersome and essentially only performed for research purposes. Therefore, the application of the deﬁnition of CKD in cirrhosis is challenging. When the serum creatinine reaches the threshold of 133 mmol/l (1.5 mg/dl), the patient is said to have type 2 HRS. The prognosis of patients with cirrhosis and CKDdwhether type 2 HRS or structural renal diseasedis worse than the corresponding stage of CKD in non-cirrhotic patients because of coexisting liver disease. Therefore, unlike non-cirrhotic patients, these patients usually do not survive long enough for the CKD to slowly deteriorate, nor will their CKD typically decline to the point of requiring dialysis unless AKI supervenes. Nevertheless, to be useful, a HRD classiﬁcation system must include all potential scenarios where CKD and advanced liver disease coexist, either as independent entities or as the result of complex organ interactions. For example, a patient with cirrhosis due to non-alcoholic steatohepatitis may also have CKD on the basis of diabetes. Similarly, a patient with cirrhosis and ascites and mild renal dysfunction below the level deﬁned by type 2 HRS may develop other forms of CKD such as IgA nephropathy related to his alcoholic liver disease. Finally, patients in both of these examples are likely to be at increased risk for AKI with various precipitants such as radiocontrast dye or sepsis. Further research is required to understand the clinical signiﬁcance of reaching K/DOQI criteria for CKD in a patient with cirrhosis. Nevertheless, the Working Group proposed the deﬁnition of CKD as an estimated GFR of <60 ml/min calculated

Table 1 The Acute Kidney Injury Network (AKIN) criteria for the definition and classification of AKI (modified RIFLE criteria)43 47 AKI stage

Serum creatinine criteria

Urine output criteria

1 (Risk)

Increase in serum creatinine of $26.4 mmol/l ($0.3 mg/dl) within 48 h or an increase of $150e200% (1.5e2-fold) from baseline Increase in serum creatinine to 200e299% (>2e3-fold) from baseline Increase in serum creatinine to $300% (>3-fold) from baseline or serum creatinine of $354 mmol/l ($4.0 mg/dl) with an acute increase of $44 mmol/l ($0.5 mg/dl) or initiation of renal replacement therapy

<0.5 ml/kg/h for >6 h

2 (Injury) 3 (Failure)

706

<0.5 ml/kg/h for >12 h <0.3 ml/kg/h for 24 h or anuria for 12 h

Gut 2011;60:702e709. doi:10.1136/gut.2010.236133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology Table 3 Definition and stages of chronic kidney disease based on kidney disease outcomes quality initiatives (K/DOQI) guidelines51 Stage

Description

GFR (ml/min/1.73 m2)

I

Kidney damage with normal or increased GFR Kidney damage with mildly decreased GFR Moderately decreased GFR Severely decreased GFR Kidney failure

$90

II III IV V

60e89 30e59 15e29 <15 (or dialysis)

Chronic kidney disease is defined as either kidney damage or glomerular filtration rate (GFR) <60 ml/min/1.73 m2 for >3 months. Kidney damage is defined as pathological abnormalities or markers of damage including abnormalities in blood or urine tests or imaging studies.

formulate a speciﬁc equation for the calculation of GFR with an acceptable accuracy for patients with advanced cirrhosis and to determine what threshold represents an increased risk in this population. The use of imaging criteria or histology to diagnose CKD was not considered by the group, as chronic renal damage may well precede the appearance of small sized kidneys on imaging and renal biopsy in patients with cirrhosis and coaguloapthy is associated with an increased risk of bleeding. Furthermore, large volume ascites may preclude this technique.42

DEFINITION OF ACUTE-ON-CHRONIC KIDNEY DISEASE IN CIRRHOSIS Figure 4 Classification of hepatorenal disorder (HRD). (A) Patients with acute liver failure without pre-existing liver disease who present with rapid worsening of renal function. (B) Spectrum of hepatorenal disorders in patients with advanced cirrhosis. AKI, acute kidney injury; CKD, chronic kidney disease; KD, kidney disease; HRS, hepatorenal syndrome. using the MDRD6 formula52 for >3 months for patients with cirrhosis (table 2) rather than a serum creatinine-based deﬁnition, to be in line with the deﬁnition of CKD in other subspecialties. The group also recognises that the MDRD6 formula tends to overestimate the GFR, but this is the formula that approximates most closely to GFR measurement using 125 I iothalamate clearance, especially in patients with low GFR of <40 ml/min.36 Future studies are therefore needed to Table 2

Proposed diagnostic criteria of kidney dysfunction in cirrhosis

Diagnosis

Definition

Acute kidney injury

Rise in serum creatinine of $50% from baseline or a rise of serum creatinine by $26.4 mmol/l ($0.3 mg/dl) in <48 h HRS type 1 is a specific form of acute kidney injury Glomerular filtration rate of <60 ml/min for >3 months calculated using MDRD6 formula HRS type 2 is a specific form of chronic kidney disease Rise in serum creatinine of $50% from baseline or a rise of serum creatinine by $26.4 mmol/l ($0.3 mg/dl) in <48 h in a patient with cirrhosis whose glomerular filtration rate is <60 ml/min for >3 months calculated using MDRD6 formula

Chronic kidney disease

Acute-on-chronic kidney disease

Both the acute deterioration in renal function and the background chronic renal dysfunction can be functional or structural in nature. HRS, hepatorenal syndrome; MDRD6, Modification of Diet in Renal Disease formula calculated using six variables of serum creatinine, age, gender, albumin, blood urea nitrogen and whether or not the patient is African-American.

Gut 2011;60:702e709. doi:10.1136/gut.2010.236133

Finally, it is important to recognise that AKI may also occur in patients with cirrhosis and pre-existing renal dysfunction. The clinician should not have any problem identifying the precipitation of type 1 HRS with spontaneous bacterial peritonitis in a patient with type 2 HRS as the deﬁnitions of types 1 and 2 HRS are well established. However, type 1 HRS may also superimpose on CKD that does not fulﬁl type 2 HRS criteria, either because the renal dysfunction is not severe enough or because it is due to other forms of kidney disease (eg, diabetic nephropathy). Under the current diagnostic criteria for HRS, this scenario poses a diagnostic dilemma as the classical deﬁnition of HRS does not permit the presence of any evidence of structural renal damage. This may have potential clinical implications as therapeutic interventions, such as a TIPS shunt, may not be inserted into patients with mixed HRD when such interventions are actually beneﬁcial.53 54 We recognise that acute-on-chronic renal failure does occur in cirrhosis, although much work is needed to understand this entity better, particularly when forms of HRD are mixed (eg, AKI superimposed on CKD in a patient with advanced liver disease). The Working Group agreed that, at present, an empirical deﬁnition of acute-on-chronic kidney disease as an increase in serum creatinine of >50% from baseline or a rise in serum creatinine of $26.4 mmol/l ($0.3 mg/dl) in <48 h in a patient with cirrhosis whose baseline GFR is <60 ml/ min calculated with the MDRD6 formula for >3 months will be adopted. Once again, both the acute deterioration in renal function and the background chronic renal dysfunction can be functional or structural in nature (table 2).

SUMMARY Renal complications are common in cirrhosis, especially in patients with refractory ascites, and they can negatively impact survival. The IAC has set out clear diagnostic criteria for both acute and chronic forms of HRS, but has not delineated guidelines for the diagnosis of other forms of renal impairment in cirrhosis, 707

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology whether acute or chronic. Well-accepted deﬁnitions and staging systems for CKD and AKI exist but have not been consistently applied to patients with advanced liver disease. Working together, the ADQI and IAC have proposed uniform standards for the diagnosis of AKI and CKD in cirrhosis adapted from these established deﬁnitions. These new diagnostic criteria are not meant to replace the well-established diagnostic criteria of HRS, but rather to broaden the scope to include other forms of renal disease in cirrhosis. The Working Group recognises that these diagnostic criteria will need to be validated in large cohorts of patients and may need to be modiﬁed depending on the outcome of these studies. Once conﬁrmed, these diagnostic criteria will be applied to the development of clinical trials to evaluate potential treatment options for patients with HRD.

RECOMMENDATIONS FOR FUTURE RESEARCH Future research will have to concentrate on improving our understanding of renal dysfunction in cirrhosis. To achieve this, the Working Group has suggested the following broad research areas. 1. To collect data on the epidemiology in terms of incidence, prevalence and basic demographics of patients with HRD including AKI, CKD and acute-on-chronic HRD. 2. To study the natural history of HRD, especially to search for precipitating factors for AKI, and to describe the rate of decline of renal function in patients with CKD. 3. To ﬁnd better markers for renal dysfunction in cirrhosis such as cystatin C or NGAL. Alternatively, to determine the test characteristics including cut-off values for serum creatinine that deﬁne an increased risk for patients with HRD. 4. To determine whether the pathophysiology of type 2 HRS is different or the same as that of type 1 HRS, but only a matter of extent. 5. To investigate whether patients with various forms of CKD are at risk of developing type 2 HRS, and to determine whether the development of type 1 HRS from pre-existing type 2 HRS is the same as from pre-existing non-functional CKD. 6. To determine novel risk factors including cardiac dysfunction for the development of renal dysfunction and their impact on prognosis. Once the various aspects of the different types of HRD are deﬁned, we will be in a better position to reﬁne the treatment for renal dysfunction in cirrhosis. The questions that will need to be answered include: 1. Are the diagnostic criteria for HRS too restrictive? Should the cut-off value of serum creatinine of 133 mmol/l (1.5 mg/dl) and 220 mmol/l (2.5 mg/dl) be lowered so that cirrhotic patients with milder renal failure can be identiﬁed and given treatment earlier? 2. Are treatments for type 1 HRS in patients with underlying CKD equally efﬁcacious as in those without? 3. In case of non-responsiveness to vasoconstrictors and albumin, how long does type 1 HRS remain a functional disease before acute tubular necrosis develops? Achieving these research goals will go a long way towards improving the outcome in patients with cirrhosis with renal dysfunction. Acknowledgements The authors wish to acknowledge the University of Southern California for the organisation of this consensus process. Funding The conferences were supported by unrestricted educational grants from Ikaria, Gambro Renal Care, Otsuka Pharmaceutical, Nx-Stage Medical, IV League Inc and Baxter Inc. 708

Competing interests None. Contributors FW searched the literature and wrote the paper. MKN organised the meeting searched the literature. All authors contributed to the content of the paper. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

Garcia-Tsao G, Parikh CR, Viola A. Acute kidney injury in cirrhosis. Hepatology 2008;48:2064e77. Kellum JA, Bellomo R, Ronco C. Acute Dialysis Quality Initiative (ADQI): methodology. Int J Artif Organs 2008;31:90e3. Flint A. Clinical report on hydroperitoneum based on analysis of 46 cases. Am J Med Sci 1863;45:306e39. Papper S, Belsky JL, Bleifer KH. Renal failure in Laennec’s cirrhosis of the liver. I. Description of clinical and laboratory features. Ann Intern Med 1959;51:759e73. Epstein M, Berk DP, Hollenberg NK, et al. Renal failure in the patient with cirrhosis. The role of active vasoconstriction. Am J Med 1970;49:175e85. Rodes J, Bosch J, Arroyo V. Clinical types and drug therapy of renal impairment in cirrhosis. Postgrad Med J 1975;51:492e7. Bartoli E, Chiandussi L, eds. Hepato-Renal Syndrome. Padua: Piccin Medical Books, 1979. Anon. Hepatorenal syndrome or hepatic nephropathy? Lancet 1980;315:801e2. Arroyo V, Gines P, Gerbes AL, et al. Definition and diagnostic criteria of refractory ascites and hepatorenal syndrome in cirrhosis. International Ascites Club. Hepatology 1996;23:164e76. Salerno F, Gerbes A, Gines P, et al. Diagnosis, prevention and treatment of hepatorenal syndrome in cirrhosis. Gut 2007;56:1310e18. Laleman W. Role of vasoactive substances and cellular effectors in the pathophysiology of cirrhotic portal hypertension: the past, the present and the futureeGeorges Brohe´e Lecture. Acta Gastroenterol Belg 2009;72:9e16. Rodrı´guez-Vilarrupla A, Ferna´ndez M, Bosch J, et al. Current concepts on the pathophysiology of portal hypertension. Ann Hepatol 2007;6:28e36. Blendis L, Wong F. The hyperdynamic circulation in cirrhosis: an overview. Pharmacol Ther 2001;89:221e31. Colle I, Geerts AM, Van Steenkiste C, et al. Hemodynamic changes in splanchnic blood vessels in portal hypertension. Anat Rec (Hoboken) 2008;291:699e713. Arroyo V, Terra C, Gines P. Advances in the pathogenesis and treatment of type-1 and type-2 hepatorenal syndrome. J Hepatol 2007;46:935e46. Cazzaniga M, Salerno F, Visentin S, et al. Increased flow-mediated vasodilation in cirrhotic patients with ascites: relationship with renal resistive index. Liver Int 2008;28:1396e401. Bernadich C, Bandi JC, Piera C, et al. Circulatory effects of graded diversion of portal blood flow to the systemic circulation in rats: role of nitric oxide. Hepatology 1997;26:262e7. Bosch J, Pizcueta MP, Fernandez M, et al. Hepatic, splanchnic and systemic haemodynamic abnormalities in portal hypertension. Baillieres Clin Gastroenterol 1992;6:425e36. Wong F, Pantea L, Sniderman K. Midodrine, octreotide, albumin, and TIPS in selected patients with cirrhosis and type 1 hepatorenal syndrome. Hepatology 2004;40:55e64. Wong F, Sniderman K, Liu P, et al. The effects of transjugular intrahepatic portasystemic shunt on systemic and renal hemodynamics and sodium homeostasis in cirrhotic patients with refractory ascites. Ann Intern Med 1995;122:816e22. Lang F, Tschernko E, Schulze E, et al. Hepatorenal reflex regulating kidney function. Hepatology 1991;14:590e4. Solis-Herruzo JA, Duran A, Favela V, et al. Effects of lumbar sympathetic block on kidney function in cirrhotic patients with hepatorenal syndrome. J Hepatol 1987;5:167e73. Badalamenti S, Graziani G, Salerno F, et al. Hepatorenal syndrome: new perspectives in pathogenesis and treatment. Arch Intern Med 1993;153:1957e67. Laffi G, La Villa G, Pinzani M, et al. Arachidonic acid derivatives and renal function in liver cirrhosis. Semin Nephrol 1997;17:530e48. Wong F. Hepatorenal syndrome. In: Lerma EV, Nissenson AR, Berns JS, eds. Current Diagnosis and Treatment in Nephrology and Hypertension. McGraw Hill: New York 2009:99e108. Stadlbauer V, Wright GA, Banaji M, et al. Relationship between activation of the sympathetic nervous system and renal blood flow autoregulation in cirrhosis. Gastroenterology 2008;134:111e19. Wong F. Cirrhotic cardiomyopathy. Hepatol Int 2009;3:294e304. Moller S, Henriksen JH. Cirrhotic cardiomyopathy: a pathophysiological review of circulatory dysfunction in liver disease. Heart 2002;87:9e15. Ruiz-del-Arbol L, Urman J, Fernandez J, et al. Systemic, renal, and hepatic hemodynamic derangement in cirrhotic patients with spontaneous bacterial peritonitis. Hepatology 2003;38:1210e18. Ruiz-del-Arbol L, Monescillo A, Arocena C, et al. Circulatory function and hepatorenal syndrome in cirrhosis. Hepatology 2005;42:439e47. Jalan R, Forrest EH, Redhead DN, et al. Reduction in renal blood flow following acute increase in the portal pressure: evidence for the existence of a hepatorenal reflex in man? Gut 1997;40:664e70.

Gut 2011;60:702e709. doi:10.1136/gut.2010.236133

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Hepatology 32. 33. 34. 35. 36. 37.

38. 39. 40. 41. 42. 43.

44. 45. 46. 47.

Krag A, Bendtsen F, Henriksen JH, et al. Low cardiac output predicts development of hepatorenal syndrome and survival in patients with cirrhosis and ascites. Gut 2010;59:105e10. Caregaro L, Menon F, Angeli P, et al. Limitations of serum creatinine level and creatinine clearance as filtration markers in cirrhosis. Arch Intern Med 1994;154:201e5. Orlando R, Floreani M, Padrini R, et al. Evaluation of measured and calculated creatinine clearances as glomerular filtration markers in different stages of liver cirrhosis. Clin Nephrol 1999;51:341e7. Sherman DS, Fish DN, Teitelbaum I. Assessing renal function in cirrhotic patients: problems and pitfalls. Am J Kidney Dis 2003;41:269e78. Gonwa TA, Jennings L, Mai ML, et al. Estimation of glomerular filtration rates before and after orthotopic liver transplantation: evaluation of current equations. Liver Transpl 2004;10:301e9. MacAulay J, Thompson K, Kiberd BA, et al. Serum creatinine in patients with advanced liver disease is of limited value for identification of moderate renal dysfunction: are the equations for estimating renal function better? Can J Gastroenterol 2006;20:521e6. Gaspari F, Perico N, Remuzzi G. Measurement of glomerular filtration rate. Kidney Int 1997;63(Suppl):S151e4. Davenport A. Difficulties in assessing renal function in patients with cirrhosis: potential impact on patient treatment. Intensive Care Med. Published Online First 4 March 2011. PMID: 21373822. Gerbes AL, Gulberg V, Bilzer M, et al. Evaluation of serum cystatin C concentration as a marker of renal function in patients with cirrhosis of the liver. Gut 2002;50:106e10. Gerbes AL, Benesic A, Vogeser M, et al. Serum NGAL: a sensitive novel marker of renal impairment in liver cirrhosis? Digestion. In press. Francoz C, Glotz D, Moreau R, et al. The evaluation of renal function and disease in patients with cirrhosis. J Hepatol 2010;52:605e13. Bellomo R, Ronco C, Kellum JA, et al. Acute renal failure e definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8:R204e12. Hoste EA, Kellum JA, Katz NM, et al. Epidemiology of acute kidney injury. Contrib Nephrol 2010;165:1e8. Cruz DN, Bagshaw SM, Ronco C, et al. Acute kidney injury: classification and staging. Contrib Nephrol 2010;164:24e32. Ricci Z, Cruz D, Ronco C. The RIFLE criteria and mortality in acute kidney injury: a systematic review. Kidney Int 2008;73:538e46. Mehta RL, Kellum JA, Shah SV, et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007;11:R31.

Gut 2011;60:702e709. doi:10.1136/gut.2010.236133

48. 49. 50. 51. 52.

53. 54.

Lassnigg A, Schmidlin D, Mouhieddine M, et al. Minimal changes of serum creatinine predict prognosis in patients after cardiothoracic surgery: a prospective cohort study. J Am Soc Nephrol 2004;15:1597e605. Jenq CC, Tsai MH, Tian YC, et al. RIFLE classification can predict short-term prognosis in critically ill cirrhotic patients. Intensive Care Med 2007;33:1921e30. Cholongitas E, Calvaruso V, Senzolo M, et al. RIFLE classification as predictive factor of mortality in patients with cirrhosis admitted to intensive care unit. J Gastroenterol Hepatol 2009;24:1639e47. National Kidney Foundation. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis 2002;39:S1e266. Levey AS, Bosch JP, Lewis JB, et al; Modification of Diet in Renal Disease Study Group. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Ann Intern Med 1999;130:461e70. Michl P, Gulberg V, Bilzer M. Transjugular intrahepatic portosystemic shunt for cirrhosis and ascites: Effects in patients with organic or functional renal failure. Scand J Gastroenterol 2000;35:654e8. Rossle M, Gerbes AL. TIPS for the treatment of refractory ascites, hepatorenal syndrome and hepatic hydrothorax: a critical update. Gut 2010;59:988e1000.

APPENDIX 1 Participants of the joint ADQIeIAC meeting Jawad Ahmad, Mount Sinai Hospital, USA; Ali Al-Khafaji, University of Pittsburgh School of Medicine, USA; Paolo Angeli, University of Padova, Italy; Vincente Arroyo, University of Barcelona, Spain; Rinaldo Bellomo, Melbourne University, Australia; Pat Brophy, University of Iowa, USA; Jorge Cerda, Albany Medical College, USA; Mink Chawla, George Washington Medical Center, USA; Andrew Davenport, University College London Medical School, UK; Connie Davis, University of Washington, USA; Yuri Genyk, University of Southern California, USA; Alexander Gerbes, Ludwig-Maximilians-Universita¨t Mu¨nchen, Germany; Noel Gibney, University of Alberta, Canada; Rajiv Jalan, University College London Medical School, UK; John A Kellum, University of Pittsburgh School of Medicine, USA; Mitra K Nadim, University of Southern California, USA; Richard Moreau, Hoˆpital Beaujon, France; Neesh Pannu, University of Alberta, Canada; Claudio Ronco, San Bortolo Hospital, Italy; Francesco Salerno, University of Milan, Italy; Randall Sung, University of Michigan, USA; Khajohn Tiranathanagul, Chulalongkorn University, Bangkok; Ashita Tolwani, University of Alabama, USA; Florence Wong, University of Toronto, Canada.

709

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

P16 reactivation induces anoikis and exhibits antitumour potency by downregulating Akt/survivin signalling in hepatocellular carcinoma cells Huanzhang Hu,1,2 Zhigang Li,3 Jie Chen,3 Duanming Wang,1,4 Juming Ma,5 Weiguo Wang,5 Jiang Li,1 Hongping Wu,1 Linfang Li,1 Mengchao Wu,1 Qijun Qian,1 Jingbo Chen,4 Changqing Su1 < Additional figures are

published online only. To view these files please visit the journal online (http://gut.bmj. com). 1

Laboratory of Viral & Gene Therapy, Eastern Hepatobiliary Surgical Hospital, The Second Military Medical University, Shanghai, China 2 Department of Hepatobiliary Surgery, Fuzhou General Hospital of Nanjing Miltary Area, Fujian, China 3 Department of Hematology & Cardiothoracic Surgery, Changhai Hospital, The Second Military Medical University, Shanghai, China 4 College of Animal Science and Technology, Shihezi University, Xinjiang, China 5 Department of Internal Medicine, 117th Hospital of Nanjing Military Area, Hangzhou, China Correspondence to Prof Changqing Su, Laboratory of Viral & Gene Therapy, Eastern Hepatobiliary Surgical Hospital, The Second Military Medical University, No. 225 Changhai Road, Shanghai 200438, China; [email protected] Published Online First 22 October 2010

ABSTRACT Hepatocellular carcinoma (HCC) is one of the most malignant tumours with high rate of recurrence and metastasis. In HCC, deficiency of the P16/CDK4/Rb pathway is a frequent molecular event, and transferring the P16 gene into cancer cells can induce cell cycle arrest and apoptosis, suggesting that the P16 gene is a good target in cancer gene therapy. The previous study demonstrated that P16 re-expression mediated by adenovirus within cancer cells can induce cell apoptosis and exert potent antitumour efficacy in cancer xenografts in nude mice. However, the molecular mechanism of P16-induced apoptosis in cancer cells is not clear yet. In this resulting study, we found that P16 re-expression can downregulate survivin expression in HCC cells. As a member of the inhibitors of the apoptotic gene family, survivin has been reported to be overexpressed in most common human cancers and present multiple physiological and pathological functions including cell cycle control, inhibition of cell apoptosis, regulation of cell division and induction of angiogenesis, etc. Further investigation found that P16 reactivation led to a decrease of phosphorylated Akt on Thr308 and phosphorylated survivin on Thr34, then downregulated survivin expression. The P16-mediated decrease of nuclear survivin in cancer cells limited CDK4 import into nuclei, which restrained CDK4 functions of promoting cell proliferation, then exhibited the effect of cell cycle arrest and induction of detachment-induced apoptosis (anoikis). The antitumor potency of P16 by downregulating the Akt/survivin signalling was also demonstrated in HCC xenograft models in nude mice. This new insight into P16 function would help in designing better strategies for cancer gene therapy.

INTRODUCTION Cell cycle progression is under the control of various interacting proteins such as cyclins, cyclindependent kinases (CDKs) and cyclin-dependent kinase inhibitors (CKIs). These factors constitute a major cascade link for cell cycle regulation, in which P16 is a negative regulator. P16 can block cell 710

cycle progression from G1 to S phase by binding to CDK4.1 2 Inactivation of P16 is a frequent molecular event in most human cancers. Reactivation of P16 by transferring the P16 gene into cancer cells can induce G1 arrest and cell apoptosis, suggesting that the P16 gene is a good target in cancer gene therapy.3 However, the molecular mechanism of P16-induced apoptosis in cancer cells is not clear yet. We previously constructed an adenovirus vector carrying the P16 gene, which expressed P16 within cancer cells, induced cancer cell apoptosis and led to a potent antitumour efﬁcacy in cancer xenografts in nude mice.4 In the resulting study reported here, we found that P16 re-expression can downregulate survivin expression in cancer cells. This phenomenon is sufﬁcient to warrant further investigation. As a member of the inhibitors of the apoptotic gene family, survivin presents multiple physiological and pathological functions including cell cycle control, inhibition of cell apoptosis, regulation of cell division and induction of angiogenesis, especially during the mitotic process, by activating multiple signalling pathways controlling tumour maintenance.5 6 Overexpression of survivin protein has been demonstrated in most common human cancers because of the activation of various signalling pathways, but not in normal differentiated adult tissues.7 8 Overexpression of survivin in cancer cells appears to be associated with resistance to chemotherapy and radiotherapy, increased tumour recurrence and shorter patient survival.9e11 Previous evidence suggested that silencing or downregulating survivin expression can induce cancer cell apoptosis and inhibit tumour growth.12 There are many strategies to downregulate survivin, including the use of antisense oligonucleotides,13 14 small interfering RNA (siRNA),15 ribozymes16 and small molecular weight inhibitors.17 In xenograft models of human cancer in nude mice, knockdown of survivin by means of various strategies not only led to cell cycle arrest, cell apoptosis and tumour growth inhibition in a dose-dependent manner, but also resulted in the enhancement of Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science the chemotherapy or radiotherapy sensitivity of cancer cells.18e20 In the current study, we investigated the molecular mechanism of P16-induced downregulation of survivin expression in cancer cells, and found that P16 re-expression can downregulate survivin expression through the inhibition of serine/threonine kinase (Akt) phosphorylation on Thr308, and consequently result in the induction of detachment-induced apoptosis, also known as anoikis,21 22 and the antitumour potency in hepatocellular carcinoma (HCC) cells.

MATERIALS AND METHODS Cell lines and adenoviruses HCC cell lines (MHCC, HepG2, SMMC-7721 and BEL-7402) and normal ﬁbroblast cell lines (MRC-5 and BJ) (American Type Culture Collection, Manassas, Virginia, USA) were cultured in media recommended by the providers. To establish the HCC cell transfectants for P16 re-expression, an adenovirus vector carrying the P16 gene, AdCMVp16, was constructed and infected into HCC cells at a multiplicity of infection (MOI) of 10 plaqueforming units (pfu)/cell as described previously.4 Cells were synchronously and similarly infected with blank vector, AdCMV, as the negative control.

Anoikis analysis The procedures for anoikis assay were as described previously.21 Cancer cells were cultured ﬁrst in 6-well adhesion plates at a density of 105 cells/well, then harvested and transferred to suspension culture in roller bottles, with a covalently bound hydrogel layer that effectively inhibits cellular attachment, at a density of 104 cells/ml. Then, at the indicated time points, cells were harvested for use.

Examination of factor expression Cells and their transfectants were harvested and the expression of the indicated factors was examined in cell lysates by western blot as previously described,23 with mouse anti-P16, rabbit anti-P16, mouse anti-ERK1/2, rabbit anti-p-ERK1/2Thr177/Thr160, mouse anti-Akt, rabbit anti-p-AktThr308, rabbit anti-PI3K p85, mouse anti-EGFR, rabbit anti-pEGFRTyr1173, rabbit anti-p-survivinThr34 (Santa Cruz Biotechnology, Santa Cruz, California, USA), rabbit anti-survivin (R&D Systems, Miles, California, USA) and mouse anti-PTEN antibodies (Maxim Biotech, Fuzhou, China). To conﬁrm further the intracellular localisation of the indicated cell cycle regulators, the nuclear and cytoplasmic extracts were obtained from the cultured cells using nuclear and cytoplasmic extraction reagents (Pierce Biotechnology, Rockford, Illinois, USA), and examined by western blot. Total RNA was extracted with TriZol reagent (Invitrogen, Carlsbad, California, USA) from HCC cells, and was used to amplify survivin mRNA with the primers P1 (59 -cggaattcaccatgggtgccccgacg-39 ) and P2 (59 -gaagatcttcaatccatggcagccag-39 ) by reverse Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

transcriptionePCR (RTePCR). Glyceraldehyde phosphate dehydrogenase (GAPDH) was used as the loading control.

Analysis of the cell cycle and anoikis Cells were cultured for the indicated time, and 106 cells were harvested and ﬁxed in cold ethanol at a concentration of 70% overnight, then incubated with propidium iodide solution at a concentration of 50 mg/ml and RNase A at a concentration of 20 mg/ml for 30 min in darkness. The cell cycle and anoikis were quantiﬁed with a ﬂow cytometer (FACS420, BD Biosciences, San Jose, California, USA). Acridine orange (AO) and ethidium bromide (EB) solutions (Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, USA) were prepared at concentrations of 100 mg/ml in 0.1 mol/l phosphate-buffered saline (PBS; pH 7.2), and mixed in the same volume. Cells were harvested and 4 ml of mixed AO/EB solution was added to 100 ml of suspended cells on slides, covered with coverslips. The cells which had undergone anoikis were counted under a ﬂuorescence microscope (Nikon TE2000E, Nikon Instruments, Shanghai, China).

Immunofluorescent confocal assay Cancer cells were harvested and 104 cells were ﬁxed in a Lab-Tek chamber (Electron Microscopy Sciences, Hatﬁeld, Pennsylvania, USA) in 4% formaldehyde for 30 min. After washing with 0.1 M PBS (pH 7.2), and incubation with the indicated primary antibodies at 48C overnight and the secondary antibodies, ﬂuorescein isothiocyanate (FITC)-conjugated antirabbit immunoglobulin G (IgG) or tetramethylrhodamine isothiocyanate (TRITC)-conjugated antimouse IgG (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), at room temperature for 1 h, cells were observed under a laser-scanning confocal microscope (Zeiss LSM510, Carl Zeiss, Germany). 4’,6-Diamidino-2-phenylindole (DAPI) was used to stain cellular nuclei.

Knockdown of gene expression by short hairpin RNA (shRNA) vectors The shRNA vectors were constructed by Genesil Biotechnology, Wuhan, China. The survivin shRNA (Sur-shRNA) of 19 nt sense DNA (gaaagtgcgccgtgccatc) targets base pairs 436e454 of the survivin gene (HSU75285), and the Akt shRNA of 19 nt sense DNA (gactacctgcactcggaga) targets base pairs 1338e1356 of the Akt1 gene (NM_005163). A mock control shRNA (CtrlshRNA, 59 -gacttcataaggcgcatgc-39 ) was synchronously constructed. Cancer cells were transfected with shRNA vectors using lipofectamine 2000 (Invitrogen, Carlsbad, California, USA). After transfection, cells were continuously cultured for the indicated time and examined for protein expression and anoikis.

Co-immunoprecipitation Cancer cells were harvested at the indicated time points. Cell lysates were prepared and the 711

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

Figure 1 Effect of P16 reactivation on cell cycle and anoikis in hepatocellular carcinoma (HCC) cells. (A) Cell lines were cultured in adhesion plates, transferred to culture in suspension in roller bottles at a density of 104 cells/ml, then infected with AdCMV-p16 at a multiplicity of infection of 10 plaque-forming units/cell. At 48 h after infection, cells with or without infection were harvested and examined for P16 expression by western blot. Glyceraldehyde phosphate dehydrogenase (GAPDH) was used as the loading control. (B) Cell cycle and anoikis of harvested cells at a concentration of 106 cells/ml were analysed by flow cytometry. Data regarding the cell cycle and anoikis of the tested cell lines are presented as the mean6SD from three independent experiments: (a) parental cells in adhesion culture; (b) parental cells in suspension culture; (c) transfectants in suspension at 48 h after infection with AdCMV; (d) transfectants in suspension at 48 h after infection with AdCMV-p16. *p<0.05; **p<0.01. (C) Anoikis was observed in the transfectants at 48 h after infection with AdCMV-p16 by acridine orange/ethidium bromide fluorescent staining (original magnification 3400), compared with the parental cells and the transfectants infected with AdCMV. The anoikis cells were counted three times under a fluorescence microscope and are shown as a percentage of the total cells counted. *p<0.05; **p<0.01. 712

Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science co-immunoprecipitation was performed as described by Shimamoto et al24 using the corresponding speciﬁc antibodies. The mouse anti-CDK4 antibody was purchased from Santa Cruz Biotechnology.

Animal models and in vivo experiments MHCC cells were subcutaneously injected into the right ﬂanks of BALB/c (nu/nu) mice (Shanghai Experimental Animal Center, Chinese Academy of Sciences, Shanghai, China), 5.03106 cells per mouse, to establish xenografts. Three weeks later, mice were separated randomly into three groups, ﬁve mice per group. Mice in the AdCMV and AdCMV-p16 groups were given ﬁve intratumoural injections every other day with a total dosage of 109 pfu of viruses per mouse, and mice in the control group were given the same volume of viral preservation solution (10 mmol/l TriseHCl pH 8.0, 2 mmol/l MgCl2, 4% sucrose). Tumour size was measured regularly, and the tumour volume was estimated with the formula a3b230.5, in which a and b represent the maximal and minimal diameters. Mice were killed 42 days later. Tumours were removed and ﬁxed in 4% formaldehyde for 4 h. Parafﬁn-embedded consecutive sections were cut for immunohistochemistry. To investigate further the capability of cancer cell inoculation to mimic cancer cell detachment and metastasis in an animal model, luciferaseexpressing MHCC cells (MHCC-Luc) infected or not with AdCMV-p16 were injected into the abdominal cavity of BALB/c (nu/nu) mice, three mice in each group, 2.53106 cells per mouse. Two weeks later, the colonisation of cancer cells in mice was monitored by animal optical in vivo imaging.

Table 1 P16 expression induces cell cycle arrest and apoptosis in suspension cultured hepatocellular carcinoma (HCC) cells Cell lines MHCC G1 Phase S Phase G2 Phase Apoptosis HepG2 G1 Phase S Phase G2 Phase Apoptosis SMMC-7721 G1 Phase S Phase G2 Phase Apoptosis BEL-7402 G1 Phase S Phase G2 Phase Apoptosis

a

b

c

d

66.82610.60 23.1569.20 10.0363.49 0.4560.08

53.04612.21 34.20611.14 12.7764.20 6.3362.16

56.6068.51 35.0069.53 8.4062.99 3.5762.63

75.41610.87* 17.5764.95* 7.0163.21 15.2865.01y

48.84613.15 36.5467.33 11.2865.88 0.5060.39

46.1964.06 40.4265.33 13.3966.09 8.4261.99

46.49610.59 44.6768.21 8.8565.75 7.3063.00

78.53615.61y 14.3167.65y 7.1663.66 19.5469.26y

56.6566.89 34.3265.34 9.0363.80 0.8360.25

47.06611.39 45.2068.02 7.7462.18 7.9962.49

43.5566.20 46.85614.52 6.2762.68 4.4960.95

72.27615.10* 16.3964.88y 11.3563.11 20.8966.26y

63.0767.82 26.10610.41 10.8363.05 1.1060.50

59.28610.02 29.3966.89 11.3264.42 7.5163.56

53.4267.85 37.13611.48 9.4562.58 7.5764.80

70.56618.23* 18.6364.81* 10.8164.36 14.7065.74*

a, parental cells in adhesion culture; b, parental cells in suspension culture; c, transfectants in suspension at 48 h after infection with AdCMV; d, transfectants in suspension at 48 h after infection with AdCMV-p16. *p<0.05. yp<0.01, vs the corresponding parental cells (b) and AdCMV-infected cells (c).

Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

Statistical analysis The experimental data were representative of three independent experiments with similar results, and were analysed statistically by two-way analysis of variance (ANOVA) at a signiﬁcance level of p<0.05 using SPSS software version 13.0 (SPSS, Chicago, Illinois, USA).

RESULTS P16 reactivation induces anoikis and cell cycle arrest in HCC cells Because P16 is a negative regulator of the cell cycle, we investigated the effect of P16 on the cell cycle and apoptosis in HCC cell adhesion and suspension cultures. At 48 h after infection by AdCMV-p16, P16 expression was positive in AdCMV-p16infected cancer cells, but negative in the parental cells (ﬁgure 1A). The AdCMV-infected cancer cells were negative for P16 expression. P16 re-expression resulted in an increase of apoptosis and an obvious cell cycle arrest in adhesion cultured cancer cells (Supplementary ﬁgure S1). However, in suspension culture, P16 re-expression induced much more apoptosis at the same MOI (ﬁgure 1B, table 1). The apoptosis which occurred in ultra-low clustered suspension culture was called anoikis. Interestingly, compared with the parental cancer cells in adhesion culture, cancer cells in suspension without infection by adenoviruses had a higher anoikis population at the beginning of suspension culture. By AO/EB staining, the anoikis rates were 1.8460.35%, 2.3660.56%, 6.8662.32% and 1.2560.22% in the parental MHCC, HepG2, SMMC-7721 and BEL-7402 cells, respectively, and 1.5160.43%, 1.1960.74%, 3.6361.52% and 2.2160.37% in the corresponding AdCMV-infected cells, respectively. When cancer cells were infected with AdCMV-p16 and cultured in suspension for 48 h, the anoikis rates were notably increased, being 15.1265.23%, 21.1767.66%, 16.3362.92% and 13.1562.67% in these tested cell lines, respectively (ﬁgure 1C).

P16 reactivation downregulates survivin expression in HCC cells Since survivin is a crucial apoptotic regulator, we investigated survivin expression in HCC cells. At 24, 48 and 72 h after infection with AdCMV-p16, P16 protein was positive in AdCMV-p16-infected HCC cells, whereas survivin protein was gradually decreased in the cell lines tested (ﬁgure 2A). However, in AdCMV-infected HCC cells, P16 was negative and survivin expression was stable as in the parental cells without infection (data not shown). RTePCR showed that survivin expression was decreased in HCC cells in suspension culture at 48 h after infection by AdCMV-p16, compared with the parental cells without infection (ﬁgure 2B). An AdCMV-p16-infected MHCC cell clone was continuously cultured to the tenth subcultured cell lines. P16 protein was positive in the cells of the ﬁrst and second subcultures, and weakly positive or negative in the cells in the ﬁfth and tenth 713

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

Figure 2 Downregulation of survivin expression in hepatocellular carcinoma (HCC) cells by P16 reactivation. (A) Cells were infected with AdCMV-p16 and harvested at the indicated time points for measurement of P16 and survivin expression by western blot, in comparison with the parental cells (Pa). The data are representative of three independent experiments with similar results. (B) Reverse transcriptionePCR (RTePCR) was used to amplify survivin mRNA from HCC cells in suspension culture after infection with AdCMV-p16 (48 h). The parental cells (Pa) were used as the control. (C) An AdCMV-p16-infected, P16-positive MHCC cell clone was cultured to the first, second, fifth and tenth subculture, and every subcultured cell line was harvested and measured for P16 and survivin expression by western blot. GAPDH, glyceraldehyde phosphate dehydrogenase.

subcultures. Correspondingly, survivin protein was expressed negatively in cells in the ﬁrst and second subcultures, and upregulated positively in the cells in the ﬁfth and tenth subcultures (ﬁgure 2C).

P16 inhibits the formation of a survivineCDK4 complex and prevents CDK4 from being transported into the nuclei in HCC cells The regulators were activated by forming a complex to regulate the cell cycle. Here, we investigated whether P16 and survivin interact with CDK4 by co-immunoprecipitation. CDK4 immunoprecipitates were positive for survivin but negative for P16 in the parental MHCC and HepG2 cells. However, when cancer cells were infected with AdCMV-p16 (48 h), the CDK4 immunoprecipitates were positive for P16, and the amount of survivin that interacts with CDK4 was notably decreased (ﬁgure 3A). The intracellular localisation of cell cycle regulators coincides with their different functions. We investigated the localisation of P16, survivin and CDK4 with the nuclear and cytoplasmic extracts of AdCMV-p16-infected MHCC cells, compared with the parental cells and AdCMV-infected cells. When the AdCMV-p16-infected MHCC cells acquired P16 re-expression mainly in cytoplasm, the nuclear survivin expression was markedly inhibited. Simultaneously, the nuclear CDK4 was decreased and cytoplasmic CDK4 was increased (ﬁgure 3B,C). 714

We further investigated the dynamic localisation of P16, survivin and CDK4 in HCC cells by confocal assay. The parental MHCC cells were negative for P16 expression, and positive for survivin protein in cellular nuclei. However, at 48 h after infection with AdCMV-p16, along with P16 expression mainly in the cytoplasm, survivin expression was notably downregulated, being weakly positive in the cytoplasm and negative in nuclei (ﬁgure 4A). In the parental MHCC cells, survivin and CDK4 were co-localised mainly in cellular nuclei. When the cancer cells were infected with AdCMV-p16 (48 h), P16 and CDK4 co-localised mainly in the cytoplasm, and the amount of nuclear survivin was notably reduced (ﬁgure 4B).

Downregulation of survivin by P16 reactivation is associated with the inhibition of phosphorylated Akt on Thr308 in HCC cells The phosphorylation of regulators is the key method for cell cycle control. We therefore investigated the phosphorylation of survivin and some regulatory factors in various signalling pathways. The parental MHCC cells were positive for survivin, epidermal growth factor receptor (EGFR), Akt and extracellular signal-regulated kinase 1/2 (ERK1/2), and nearly all the survivin was phosphorylated on Thr34 (p-survivinThr34). When cancer cells re-expressed P16, p-survivinThr34 and total survivin were notably decreased, as was Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

Figure 3 P16 and survivin interactions with cyclin-dependent kinase 4 (CDK4). (A) Cancer cell transfectants infected by AdCMV-p16 were harvested from suspension culture and examined for P16 or survivin co-immunoprecipitated with CDK4. The parental cells (Pa) were used as the control. (B) Western blot was performed to confirm the intracellular localisation of P16, survivin and CDK4 by using the nuclear and cytoplasmic extracts of AdCMV-p16-infected MHCC cells. The parental cells and AdCMV-infected cells were used as the controls. (C) Densitometric analysis was performed to show the expression levels of survivin and CDK4 in the nuclear and cytoplasmic extracts of MHCC cells, normalised with glyceraldehyde phosphate dehydrogenase (GAPDH) density. Columns are the mean of three separate experiments; bars¼SD; *p<0.05; **p<0.01. phosphorylated AktThr308 (p-AktThr308). The upstream kinase PI3K (phosphatidylinositol 3-kinase) and factor PTEN (phosphatase and tensin homologue deleted on chromosome 10) are the positive and negative regulators, respectively, that control the Akt phosphorylation level. Therefore, we examined the expression of PI3K and PTEN. The results showed that P16 re-expression in MHCC cells did not alter PI3K and PTEN expression (ﬁgure 5A). What the function of p-AktThr308 is in P16 reactivation and survivin downregulation has not been reported. When P16-deﬁcient MHCC cells were cultured in suspension and transfected with the Akt-shRNA vector, both total Akt (t-Akt) and p-AktThr308 were notably reduced, as was total survivin and p-survivinThr34, but this did not occur in MHCC cells transfected with the mock control shRNA (ﬁgure 5B). When MHCC cells were Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

transfected with survivin shRNA vector, at 24, 48 and 72 h after transfection, survivin expression was notably downregulated; however, this was not found in cells transfected with the mock control shRNA. The silencing of survivin expression did not inﬂuence p-AktThr308 levels. Correspondingly, the anoikis rates were increased in survivin shRNAtransfected cells, being 11.6764.33%, 20.9162.01% and 22.3563.67% at 24, 48 and 72 h after transfection, respectively, compared with 3.6461.67% in the parental MHCC cells. However, there was no difference between MHCC cells transfected with and without the mock control shRNA (ﬁgure 5C). As P16 exerts its function through inhibiting CDK4 kinase activity, we synchronously examined the intracellular localisation of CDK4 associated with p-AktThr308. Confocal assay showed that p-AktThr308 was positive in the cytoplasm, and most CDK4 localised in nuclei of the parental 715

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

Figure 4 Intracellular localisation of P16 and survivin interacting with cyclin-dependent kinase (CDK4). MHCC cells in a Lab-Tek chamber at a density of 104 cells/well were used to observe the expression of P16 and survivin at 48 h after infection with AdCMV-p16 (A), and the intracellular localisation of P16 or survivin interacting with CDK4 (B) by immunofluorescent confocal assay. The parental cells without infection were used as the control; original magnification 3400. DAPI, 4’,6-diamidino-2-phenylindole. 716

Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

Figure 5 Signalling pathway associated with P16-mediated downregulation of survivin expression in hepatocellular carcinoma (HCC) cells. (A) MHCC cell transfectants infected with AdCMV-p16 (48 h) were harvested to measure the expression of signal factors associated with cell cycle regulation by western blot. Pa, parental cells; BJ, normal fibroblast cell line. (B) MHCC cells were transfected with the Akt-shRNA vector, and the expression of t-Akt and p-AktThr308, total survivin and p-survivinThr34 was examined by western blot, in comparison with the parental cells (Pa). The mock control short hairpin RNA (shRNA) vector (Ctrl-shRNA) was used as the negative control. (C) MHCC cells were transfected with survivin shRNA vector (Sur-shRNA), and the expression of survivin and p-AktThr308 was examined at 24, 48 and 72 h after transfection, in comparison with the parental cells (Pa). The mock control shRNA vector (Ctrl-shRNA) was used as the negative control. Synchronously, the changes in cell anoikis associated with the silencing of survivin expression were examined by acridine orange/ethidium bromide fluorescent staining. The cell anoikis rates were counted three times under the fluorescent microscope and are shown as a percentage of the total cells counted, **p<0.01. (D) A confocal assay was used to show the intracellular transportation of cyclin-dependent kinase 4 (CDK4) from the nuclei to the cytoplasm in Akt-shRNA-transfected MHCC cells, compared with the parental cells; original magnification 3400. DAPI, 4’,6-diamidino-2-phenylindole.

Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

717

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science MHCC cells. However, after transfection with AktshRNA, p-AktThr308 was downregulated and most CDK4 existed in the cytoplasm (ﬁgure 5D).

P16 reactivation exhibits antitumour potency in HCC xenografts in nude mice In MHCC xenografts in nude mice, suppression of tumour growth in the AdCMV-p16-treated group was more effective, with a tumour inhibition rate of 73.18%, compared with the control group (ﬁgure 6A). The xenograft tumours were examined immunohistochemically. In the AdCMV and control groups, cancer cells were negative for P16 expression. Survivin was positive and mainly distributed in cell nuclei, with the percentages of 80.80612.07% and 75.60617.98%, respectively, and total Akt and p-Akt were positive in the cytoplasm and nuclei. However, in the AdCMVp16-treated group, cancer cells were positive for P16 expression, and survivin was markedly decreased, with a percentage of 16.2069.47%; p-Akt was also notably depressedy (ﬁgure 6B,C). We mimicked the in vivo process of cancer cell detachment and metastasis by inoculating MHCCLuc cells with or without infection by AdCMV-p16 into mouse abdominal cavities; the colonisation was monitored by bioluminescent imaging. Two weeks after inoculation, the bioluminescent intensity in mice inoculated with the AdCMV-p16infected cells was clearly lower than that in mice inoculated with the parental cells (ﬁgure 6D).

DISCUSSION Deﬁciency of the P16/CDK4/retinoblastoma (Rb) pathway in cancer cells can promote cell cycle progression by activating CDK4 kinase activity, inactivating Rb protein and releasing E2F transcriptional factor.25 In our previous study, reactivation of P16 function in P16-deﬁcient cancer cells induced programmed cell death (PCD) and showed antitumour efﬁcacy.4 PCD is a tightly regulated type of cell death. There are several distinct subtypes of PCD that can be triggered through the intrinsic or extrinsic pathway.26 Although these subtypes of PCD are similar in morphology, they exert different functions in different physiological processes. Apoptosis induced by loss of the attachment to the substrate or to other cells is called anoikis.27 Anoikis acts as a functional barrier to cancer cell spread and metastasis. When cancer cells acquire resistance to anoikis, they may present an aggressive nature and have survival ability in the systemic circulation, thereby facilitating secondary tumour formation in distant organs. Therefore, anoikis is a major regulatory factor in the recurrence and metastasis of cancers.28 To investigate further the molecular mechanisms of P16-induced anoikis, this study found that P16 reactivation mediated by AdCMV-p16 led to cell cycle arrest and downregulation of survivin expression in HCC cells. To date, the mechanisms of anoikis involved in P16 and survivin, and their interactions in cancer cells, have not been clearly deﬁned. P16 and survivin are a pair of incompatible 718

factors in the cell cycle and cell apoptosis regulation. Along with P16 reactivation, the anoikis rates of HCC cells were notably increased. When cancer cells were cultured for a long time to the ﬁfth or tenth subculture, the adenovirus-mediated P16 expression was gradually lost and survivin expression was re-upregulated, demonstrating that downregulation of survivin expression is due to the reactivation of P16 function. CDK4 is the key factor in cell cycle control. Cyclin D1 and P16 eventually inﬂuence the G1eS checkpoint by interacting with CDK4.29 By coimmunoprecipitation, western blot using the nuclear and cytoplasmic extracts and immunolabelling confocal assays, we found that CDK4 was co-immunoprecipitated with survivin in P16-deﬁcient HCC cells, both co-localised mainly in cancer cell nuclei. However, when HCC cells re-expressed P16, the P16 protein competed with survivin for CDK4 binding and transported CDK4 from the nuclei to the cytoplasm. The results were consistent with previous ﬁndings.30 The translocation of CDK4 from the nuclei to the cytoplasm highlighted that its function to accelerate cell proliferation was weakened. Interestingly, we found an important molecular eventdthat is, that the nuclear survivin was notably reduced in HCC cells. It was reported that survivin in cancer cells exerts different functions depending on its localisation. The nuclear survivin is involved in promoting cell proliferation, whereas the cytoplasmic survivin may participate in controlling cell survival.31e33 Accordingly, survivin promotes cell proliferation and inhibits cell apoptosis by competitively interacting with CDK4, and the P16-induced decrease in the amount of survivin and CDK4 in nuclei contributes to the inhibition of cell cycle progression and induction of cell anoikis in cancer cells. Cell cycle progression requires the sequential activation of a series of phosphorylated CDKs. P16 inhibits the kinase activity of CDK4 in competition with cyclin D1.25 We found that most survivin existing in HCC cells was phosphorylated on Thr34. When cancer cells re-expressed P16, the phosphorylated survivin as well as the total survivin protein was notably depressed. This process also inﬂuenced the phosphorylation of Akt on Thr308, but not the phosphorylation of EGFR and ERK, demonstrating that the P16-mediated downregulation of survivin phosphorylation and induction of cell anoikis may be through the phosphorylated Akt pathway. Moreover, not only was the antiparallel relationship between P16 and survivin expression levels, or between P16 expression and the Akt phosphorylation level, found in HCC cell lines, but it was also proved to occur in the primary HCC specimens (Supplementary ﬁgure S2A,B). To investigate further the effect of Akt signalling between the kinase activity of CDK4 and the phosphorylation of survivin, we silenced Akt and survivin expression using their targeting shRNA vectors. When the total Akt and phosphorylated Akt were silenced, total survivin and phosphorylated survivin were decreased, and CDK4 was Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

Figure 6 Antitumour potency of P16 in hepatocellular carcinoma (HCC) xenografts. (A) HCC xenograft models were established by subcutaneous injection of 5.03106 MHCC cells per mouse in three groups of nude mice (n¼5/group). After repeated intratumoural administration of adenoviruses, tumour growth was observed regularly. The antitumour potency of AdCMV-p16 was more effective compared with the control group and the AdCMV group; **p<0.01 vs the control group (left panel) for comparison of tumour size among the three groups (right panel). (B) Expression of P16, survivin, t-Akt and p-Akt in xenograft tumours was examined immunohistochemically; original magnification 3300. (C) Survivin, t-Akt and p-Akt expression in xenograft tumours was quantified as a percentage within five high-power fields under a microscope and is shown in histograms; *p<0.05, **p<0.01. (D) In vivo cancer cell detachment and metastasis was mimicked in a mouse model. A total of 2.53106 MHCC-Luc cells with or without infection by AdCMV-p16 were injected into the abdominal cavities of nude mice. Bioluminescent imaging was performed 2 weeks later to monitor the abdominal colonisation. The bioluminescent intensities in mice bearing the AdCMV-p16-infected MHCC-Luc cell tumours or the parental MHCC-Luc cell tumours are shown. Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

719

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science transported from the nuclei to the cytoplasm in cancer cells. The silencing of survivin expression could increase cancer cell anoikis, but did not inﬂuence the levels of phosphorylated Akt, demonstrating that survivin is one of the target substrates of Akt. It seems that the inhibition of Akt kinase activity can depress the phosphorylation of survivin and cause CDK4 to be exported from the nuclei. The phosphorylated Akt and survivin may be necessary for the dynamic localisation of CDK4. A previous study showed that the inhibition of survivin phosphorylation on Thr34 resulted in loss of survivin expression, and the dephosphorylated survivin exhibited an accelerated clearance as compared with the phosphorylated survivin,34 demonstrating that the active form of phosphorylated survivin is required to maintain survivin expression and stability in cancer cells. Once P16 function was reactivated in HCC cells, the CDK4 kinase activity was inhibited, the nuclear survivin was released by P16 competition by the survivineCDK4 complex, and the released and dephosphorylated survivin was not only unstabilised and easily degradable, but also incapable of maintaining expression. Therefore, we supposed that this is one of the possible reasons for the decrease in the nuclear survivin in HCC cells re-expressing P16. Although many factors and many pathways are involved in cell cycle regulation, strong evidence demonstrated that the cell cycle progression is mainly regulated by the pathway consisting of cyclins, CDKs and CKIs.35 Moreover, activation of Akt signalling plays a pivotal role in fundamental cellular functions such as the antiapoptotic effect by phosphorylating a variety of substrates,36 37 including survivin. We found that P16-mediated inhibition of Akt/survivin signalling induced cell cycle arrest and anoikis in HCC cells, and ﬁnally exerted antitumour potency, inhibiting cancer dissemination in HCC xenograft models in nude mice. To investigate the regulatory pathway between P16 expression and Akt inactivation, we examined the expression of an upstream kinase and regulator that control the Akt phosphorylation level. The results showed that P16-mediated downregulation of phosphorylated Akt did not involve PI3K and PTEN. Several reports also conﬁrmed that Akt activity could be regulated in a PI3K-independent manner.38 The upstream regulatory molecules involved in regulation of Akt signalling are numerousdfor example, insulin-like molecules, growth factor 1 receptor and some Aktinteracting proteins.39 Therefore, more work needs to be done regarding this regulatory pathway. We have provided the ﬁrst evidence that P16 reactivation induces anoikis and antitumour potency by downregulating Akt/survivin signalling in HCC, which will help us to design better strategies for cancer gene therapy. Funding This work was supported by grants from the National Significant Science and Technology Special Projects of New Drugs Creation (No. 2009ZX09102-235) and the National Natural Scientific Foundation of China (No. 81071866, 30973469). 720

Competing interests None. Patient consent Obtained. Ethics approval This study was conducted with the approval of the Research Ethics Committee of the Second Military Medical University. Contributors HH, ZL and JC contributed equally to this work. Provenance and peer review Not commissioned; externally peer reviewed.

REFERENCES 1. 2. 3. 4.

5. 6. 7. 8. 9. 10. 11. 12. 13.

14. 15.

16. 17.

18.

19.

20. 21. 22. 23.

Ca´nepa ET, Scassa ME, Ceruti JM, et al. INK4 proteins, a family of mammalian CDK inhibitors with novel biological functions. IUBMB Life 2007;59:419e26. Coqueret O. Linking cyclins to transcriptional control. Gene 2002;299:35e55. Allay JA, Steiner MS, Zhang Y, et al. Adenovirus p16 gene therapy for prostate cancer. World J Urol 2000;18:111e20. Ma J, He X, Wang W, et al. E2F promoter-regulated oncolytic adenovirus with p16 gene induces cell apoptosis and exerts antitumour effect on gastric cancer. Dig Dis Sci 2009;54:1425e31. Ryan BM, O’Donovan N, Duffy MJ. Survivin: a new target for anticancer therapy. Cancer Treat Rev 2009;35:553e62. Sah NK, Khan Z, Khan GJ, et al. Structural, functional and therapeutic biology of survivin. Cancer Lett 2006;244:164e71. Andersen MH, Svane IM, Becker JC, et al. The universal character of the tumour-associated antigen survivin. Clin Cancer Res 2007;13:5991e4. Yamashita S, Masuda Y, Kurizaki T, et al. Survivin expression predicts early recurrence in early-stage breast cancer. Anticancer Res 2007;27:2803e8. Pennati M, Folini M, Zaffaroni N. Targeting survivin in cancer therapy: fulfilled promises and open questions. Carcinogenesis 2007;28:1133e9. Johnson ME, Howerth EW. Survivin: a bifunctional inhibitor of apoptosis protein. Vet Pathol 2004;41:599e607. Cheung CH, Chen HH, Kuo CC, et al. Survivin counteracts the therapeutic effect of microtubule de-stabilizers by stabilizing tubulin polymers. Mol Cancer 2009;8:43. Zaffaroni N, Pennati M, Daidone MG. Survivin as a target for new anticancer interventions. J Cell Mol Med 2005;9:360e72. Ro¨del F, Frey B, Leitmann W, et al. Survivin antisense oligonucleotides effectively radiosensitize colorectal cancer cells in both tissue culture and murine xenograft models. Int J Radiat Oncol Biol Phys 2008;71:247e55. Hansen JB, Fisker N, Westergaard M, et al. SPC3042: a proapoptotic survivin inhibitor. Mol Cancer Ther 2008;7:2736e45. Yang CT, Li JM, Weng HH, et al. Adenovirus-mediated transfer of siRNA against survivin enhances the radiosensitivity of human non-small cell lung cancer cells. Cancer Gene Ther 2010;17: 120e30. Zaffaroni N, Pennati M, Folini M. Validation of telomerase and survivin as anticancer therapeutic targets using ribozymes and small-interfering RNAs. Methods Mol Biol 2007;361:239e63. Giaccone G, Zatloukal P, Roubec J, et al. Multicenter phase II trial of YM155, a small-molecule suppressor of survivin, in patients with advanced, refractory, non-small-cell lung cancer. J Clin Oncol 2009;27:4481e6. Sun Y, Lin R, Dai J, et al. Suppression of tumour growth using antisense oligonucleotide against survivin in an orthotopic transplant model of human hepatocellular carcinoma in nude mice. Oligonucleotides 2006;16:365e74. Kojima H, Iida M, Yaguchi Y, et al. Enhancement of cisplatin sensitivity in squamous cell carcinoma of the head and neck transfected with a survivin antisense gene. Arch Otolaryngol Head Neck Surg 2006;132:682e5. Pennati M, Folini M, Zaffaroni N. Targeting survivin in cancer therapy. Expert Opin Ther Targets 2008;12:463e76. Douma S, Van Laar T, Zevenhoven J, et al. Suppression of anoikis and induction of metastasis by the neurotrophic receptor TrkB. Nature 2004;430:1034e9. He X, Ota T, Liu P, et al. Downregulation of HtrA1 promotes resistance to anoikis and peritoneal dissemination of ovarian cancer cells. Cancer Res 2010;70:3109e18. Su C, Peng L, Sham J, et al. Immune gene-viral therapy with triplex efficacy mediated by oncolytic adenovirus carrying an interferon-gamma gene yields efficient antitumour activity in

Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in basic science

24.

25. 26. 27. 28. 29. 30. 31.

immunodeficient and immunocompetent mice. Mol Ther 2006;13:918e27. Shimamoto S, Takata M, Tokuda M, et al. Interactions of S100A2 and S100A6 with the tetratricopeptide repeat proteins, Hsp90/ Hsp70-organizing protein and kinesin light chain. J Biol Chem 2008;283:28246e58. Mus‚at M, Vax VV, Borboli N, et al. Cell cycle dysregulation in pituitary oncogenesis. Front Horm Res 2004;32:34e62. Kroemer G, Galluzzi L, Vandenabeele P, et al. Classification of cell death: recommendations of the Nomenclature Committee on Cell Death 2009. Cell Death Differ 2009;16:3e11. Gilmore AP. Anoikis. Cell Death Differ 2005;12:1473e7. Ohtani N, Yamakoshi K, Takahashi A, et al. The p16INK4aeRB pathway: molecular link between cellular senescence and tumour suppression. J Med Invest 2004;51:146e53. Ai MD, Li LL, Zhao XR, et al. Regulation of survivin and CDK4 by EpsteineBarr virus encoded latent membrane protein 1 in nasopharyngeal carcinoma cell lines. Cell Res 2005;15:777e84. Suzuki A, Hayashida M, Ito T, et al. Survivin initiates cell cycle entry by the competitive interaction with Cdk4/p16(INK4a) and Cdk2/cyclin E complex activation. Oncogene 2000;19:3225e34. Li F, Yang J, Ramnath N, et al. Nuclear or cytoplasmic expression of survivin: what is the significance? Int J Cancer 2005;114:509e12.

32. 33. 34. 35.

36. 37. 38. 39.

Connell CM, Wheatley SP, McNeish IA. Nuclear survivin abrogates multiple cell cycle checkpoints and enhances viral oncolysis. Cancer Res 2008;68:7923e31. Liu F, Matsuura I. Inhibition of Smad antiproliferative function by CDK phosphorylation. Cell Cycle 2005;4:63e6. Wall NR, O’Connor DS, Plescia J, et al. Suppression of survivin phosphorylation on Thr34 by flavopiridol enhances tumour cell apoptosis. Cancer Res 2003;63:230e5. Schrage YM, Lam S, Jochemsen AG, et al. Central chondrosarcoma progression is associated with pRb pathway alterations: CDK4 down-regulation and p16 overexpression inhibit cell growth in vitro. J Cell Mol Med 2009;13:2843e52. Ripka S, Neesse A, Riedel J, et al. CUX1: target of Akt signalling and mediator of resistance to apoptosis in pancreatic cancer. Gut 2010;59:1101e10. Nicholson KM, Anderson NG. The protein kinase B/Akt signalling pathway in human malignancy. Cell Signal 2002;14:381e95. Filippa N, Sable CL, Filloux C, et al. Mechanism of protein kinase B activation by cyclic AMP-dependent protein kinase. Mol Cell Biol 1999;19:4989e5000. Pekarsky Y, Koval A, Hallas C, et al. Tcl1 enhances Akt kinase activity and mediates its nuclear translocation. Proc Natl Acad Sci USA 2000;97:3028e33.

Information in a hurry... If you need the latest information in emergency care then you need the Emergency Medicine Journal. Packed with research, educational papers and debate of all aspects of emergence medicine, the journal will make sure you know everything you need to. FOR MORE DETAILS OR TO SUBSCRIBE, VISIT THE WEBSITE TODAY

emj.bmj.com

Gut 2011;60:710e721. doi:10.1136/gut.2010.220020

721

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice

Chronic gastrointestinal ischaemia: shifting paradigms Peter B F Mensink,1 Leon M G Moons,1 Ernst J Kuipers1,2 1

Department of Gastroenterology and Hepatology, Erasmus MC University Medical Center, Rotterdam, The Netherlands 2 Internal Medicine, Erasmus MC University Medical Center, Rotterdam, The Netherlands Correspondence to Dr Peter B F Mensink, Department of Gastroenterology and Hepatology, Erasmus MC University Medical Center, ’s Gravendijkwal 230, 3015 CE Rotterdam, The Netherlands; [email protected] Published Online First 29 November 2010

ABSTRACT Chronic gastrointestinal ischaemia (CGI) is generally considered to be a rare disease entity. The majority of patients with CGI are only diagnosed after a long period of slowly progressive abdominal symptoms, in some cases with impressive weight loss. These patients may have a broad range of clinical signs and quite often undergo repeated extensive evaluation of their symptoms with negative outcome. The classical triad of symptoms, also known as ‘abdominal angina’, is defined as the combination of postprandial pain, weight loss due to fear of pain after eating, and an abdominal bruit during physical examination. Recent studies have shed new lights on these long unchallenged concepts. These studies first showed that CGI is more prevalent than previously thought and can occur in patients with both single- and multi-vessel disease. Second, the disease presents with a much wider range in symptoms, and only a minority of patients present with the classical triad. Third, long-term positive outcomes can be achieved after endovascular or surgical revascularisation therapy in large proportion of patients. This knowledge results from a combination of clinical research by dedicated focus groups, the current widespread availability of new imaging techniques such as CT-angiography, the development of new functional tests for assessment of mucosal perfusion, and the evolution of endovascular stenting options. Clinicians diagnosing and treating patients with acute and chronic abdominal conditions have to be aware of these new developments. We therefore here review the new insights on CGI with a focus on epidemiology, pathophysiology, current diagnostics and treatment.

INTRODUCTION Gastrointestinal ischaemia is a well known cause of abdominal symptoms. Gastroenterologists, surgeons, general physicians and intensive care specialists are all confronted regularly with patients who are clinically suspected of this potentially lifethreatening condition, in particular in the setting of an acute onset of abdominal symptoms in combination with a known medical history of vascular disease.1 The incidence of this acute condition is estimated at 13/100 000 person-years. Its management ﬁrst and foremost depends on a timely establishment of a correct diagnosis, something which often remains a clinical challenge.2 The clinical presentation of gastrointestinal ischaemia depends on the localisation, aetiological background, and speed of onset. One well-known patient category is those with acute, or sometimes chronic, lower gastrointestinal ischaemia, referred 722

to as ischaemic colitis. This can occur after surgery such as for abdominal aorta aneurysm, but the main cause of ischaemic colitis is non-occlusive mesenteric ischaemia (NOMI). This condition is usually evoked by another underlying disease such as severe anaemia, shock, and low cardiac output secondary to arrhythmias, or myocardial infarction.3e5 In contrast, chronic gastrointestinal ischaemia (CGI) is much less known and generally considered to be very rare. Patients with this condition mostly present as outpatients with chronic, slowly progressive abdominal symptoms. These patients may have a broad range of clinical signs and quite often undergo repeated extensive evaluation of their symptoms with negative outcome.6 7 The classical triad of symptoms, also known as ‘abdominal angina’, is deﬁned as the combination of postprandial pain, weight loss due to fear of pain after eating, and an abdominal bruit during physical examination. However, recent research in patients with CGI has shown that this classical triad is lacking in the majority of these patients.6 Furthermore, the generally accepted ‘rule’ that CGI only occurs in patients with signiﬁcant stenoses of at least two of the three major arteries of the mesenteric system has to be adjusted.6 Recent studies have shown that patients with single gastrointestinal arterial stenosis in combination with insufﬁcient collateral circulation can also develop CGI and can beneﬁt from revascularisation therapy.6e8 This knowledge results from a combination of clinical research by dedicated focus groups, the current widespread availability of new imaging techniques such as CTeangiography, the development of new functional tests for assessment of mucosal perfusion, and the evolution of endovascular stenting options. Clinicians diagnosing and treating patients with acute and chronic abdominal conditions have to be aware of these new developments. We therefore here review the new insights on CGI with a focus on epidemiology, pathophysiology, current diagnostics and treatment.

ANATOMY AND AETIOLOGY OF GASTROINTESTINAL ARTERIAL STENOSIS Anatomy, physiology and aetiology Three main aortic branches provide the arterial blood supply to the gastrointestinal tract: the coeliac artery (CA), superior mesenteric artery (SMA) and inferior mesenteric artery (IMA). The general gastrointestinal vascular anatomy dictates that the CA supplies stomach, liver, part of the pancreas, and the proximal part of the duodenum. Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice The SMA supplies the distal part of the duodenum, the entire small bowel and the proximal colon, and the IMA is relatively small and supplies the distal colon; see ﬁgure 1. Branches of these main gastrointestinal arteries enter the serosa of the gut on the mesenteric side to form a serosal vascular plexus around the intestine. A submucosal plexus is formed after penetration of the intestinal wall. From here, arterioles penetrate the muscularis mucosa to the superﬁcial mucosal layers. At the mucosal tip they branch into an intense capillary network of capillaries and venules. In a fasting state under basal conditions, approximately 20% of the cardiac output ﬂows through the gastrointestinal arteries. The CA receives 800 ml blood/ min, dividing this ﬂow into 350 ml/min to both the splenic and hepatic arteries. The remaining 100 ml/min supplies the stomach, duodenum and part of the pancreas. Postprandially, the ﬂow in the CA increases by 30% to 1100 ml/min. The SMA, which has a similar diameter to that of the CA, has a basal ﬂow of 500 ml/min, which increases by more than 150% after a meal, with volumes up to 1400 ml/min. The IMA is relatively small in diameter as compared to the CA and SMA; see ﬁgure 1.

The gastrointestinal arterial circulation is provided with an abundant collateral network. It is generally assumed that this collateral circulation of the arterial mesenteric tract prevents clinically relevant gastrointestinal ischaemia in the majority of cases with single- or multi-vessel gastrointestinal arterial stenosis. The aetiology of the majority of stenoses of the gastrointestinal arteries can be divided into concentric and eccentric diseases; see box 1. The coeliac artery compression syndrome (CACS) is the primary cause of gastrointestinal arterial stenosis in younger patients. In elderly patients, atherosclerotic disease is the major cause of gastrointestinal arterial stenosis. Non-concentric intra-arterial disease due to arterial thrombosis or dissection causes signiﬁcant stenoses in a minority of cases.

Epidemiology of gastrointestinal arterial stenosis Autopsy studies have provided data about the prevalence of, mostly, asymptomatic gastrointestinal arterial stenosis. Critical narrowing, deﬁned as a luminal occlusion of >50%, of either the CA, SMA or IMA was frequently found, with a prevalence ranging from 11 to 21% in two autopsy series with in total 313 subjects included, with age of

Mammary artery 700 ml/min Po r t a l v e in

Celiac artery 800→1100 ml/min 100 ml/min Bűhler anastomosis

500→1400 ml/min

Radix mesenterica

Sup. mesenteric artery

Superior mesenteric vein in f eri or mesenteric v ei n

Aorta

Arc of Riolan

Inf. mesenteric artery 50→ ? ml/min Internal iliac artery

Figure 1 Overview of the blood circulation to and from the digestive tract. The blood is supplied to the digestive tract through the coeliac artery, superior mesenteric artery and inferior mesenteric artery. There are some important anastomoses of the mesenteric arteries, which can compensate the blood supply in case of stenosis or occlusion of one of the arteries. (1) Collateral circulation from the mammary artery to the coeliac artery, (2) an anastomosis between the superior pancreaticoduodenal artery from the coeliac artery, and the inferior pancreaticoduodenal artery from the superior mesenteric artery (anastomosis of Bu¨hler), (3) an anastomosis of the superior mesenteric artery and inferior mesenteric artery (arc of Riolan or meandering mesenteric artery) and (4) an anastomosis between the superior rectal artery of the inferior mesenteric artery and the rectal arteries from the internal iliac artery. Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

723

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice

Box 1 Aetiology of gastrointestinal arterial occlusive disease Concentric:

< atherosclerotic stenosis < vasculitis < (familial) fibromuscular dysplasia

Eccentric: < coeliac artery compression syndrome (CACS) Intra-arterial: < thrombo-embolism < dissection death ranging from 19 to 97 years.9 10 A more recent Finnish autopsy study, including 120 subjects, reported a 29% prevalence of stenosis of either the CA or SMA, with the CA being the most common affected site.11 In 15% of all cases at least two gastrointestinal arteries were affected. The latter study also showed a strong correlation between gastrointestinal arterial stenosis and ageing: in subjects younger than 40 years the prevalence of gastrointestinal arterial stenosis was 6%, rising to 14% between the age 40 and 59, and to 67% in subjects aged 80 years or more.11 A large study reviewing abdominal angiographies of 713 patients showed stenosis or occlusion of at least one of the gastrointestinal arteries in 17% of cases.12 Recent studies using modern imaging modalities of the gastrointestinal arteries also show a high prevalence of, often asymptomatic, gastrointestinal arterial stenosis. In a prospective series of asymptomatic patients undergoing gastrointestinal angiography for chemoembolisation of hepatic tumours, 7% of patients were found to have a signiﬁcant stenosis of the CA.13 In a recent study in a large cohort of asymptomatic elderly Americans with a mean age of 77 years, the prevalence of stenosis in the CA and SMA was evaluated using abdominal duplex ultrasound measurements. In 18% of these subjects, a stenosis of the CA and/or SMA was found. Eighty-six per cent of these stenoses were isolated CA stenoses, 2% isolated CA occlusions, 5% isolated SMA stenoses, and 7% combined CA and SMA stenoses.14 In another report using abdominal duplex ultrasound measurements in asymptomatic subjects, a 3% prevalence of CA stenoses was reported in subjects <65 years of age, rising to 18% in subjects >65 years of age. Among patients with an isolated stenosis, 81% of lesions were present in the CA, and 19% in the SMA.15 Taken together, the latter studies suggest that the prevalence of asymptomatic gastrointestinal arterial stenosis in the general population ranges between 6 to 29% and increases with age. The exact prevalence of symptomatic gastrointestinal arterial disease is currently unknown. This is due to a variety of reasons including lack of symptom speciﬁcity, limited awareness of the spectrum of disease and the persistence of old theorems on symptoms and the prerequisite of 724

multi-vessel disease. Furthermore, the former need for invasive testing, the lack of functional tests, and the invasive character of current therapies also interfere in this process. Nevertheless, new data will provide actual insights into the prevalence of symptomatic gastrointestinal arterial disease.

SINGLE-VESSEL GASTROINTESTINAL ARTERY STENOSIS AND CGI Until recently, controversy existed about the mere existence of single-vessel CGI. It was generally thought that CGI is caused by signiﬁcant stenoses involving at least two of the three main mesenteric arteries. Consequently, it was thought that due to the supposedly abundant collateral circulation, single gastrointestinal artery stenosis, whether caused by atherosclerosis or external compression, would not give rise to clinically relevant gastrointestinal ischaemia. Adequate multimodality assessment of larger groups of patients has, however, convincingly shown that single artery stenosis can give rise to CGI. In a large cohort of patients with gastrointestinal artery stenoses, CGI was diagnosed in approximately 60% of patients with single artery disease. The most common cause of single gastrointestinal artery stenosis in this cohort was compression of the CA by the median arcuate ligament. This was the cause of 65% of the isolated CA stenoses.8 A comparison with other data is hampered by the fact that most studies do not specify the cause and character of isolated CA stenoses, being concentric or eccentric.

Coeliac artery compression syndrome as a cause of CGI The main cause of single artery stenosis is extrinsic compression due to CACS, also known as median arcuate ligament syndrome, or Dunbar’s syndrome. This condition was ﬁrst described by Lipshutz in 1917, followed decades later by Harjola (in 1963) and Dunbar et al (in 1965).16e18 The median arcuate ligament usually passes over the aorta at the level of the ﬁrst lumbar vertebral body, superior to the origin of the coeliac artery (see ﬁgure 2A). The actual cause of clinical symptoms in the CACS is still discussed. Currently, two main theories are used to explain the pathogenesis of symptoms. According to the ﬁrst hypothesis, an anomalous ﬁbrous diaphragmatic band compresses the coeliac axis in patients with a relatively low insertion of the diaphragm, especially during expiration with partial relief during inspiration. This extrinsic compression limits the inﬂow of blood, causing mucosal ischaemia. According to the second theory, the pain results from stimulation of the coeliac plexus either by mesenteric vasoconstriction or local irritation.19 CACS is typically diagnosed in women under the age of 50 years. A signiﬁcant compression of the CA by the median arcuate ligament of the diaphragm can give rise to symptoms of CGI (see ﬁgure 2B). In a recent prospective study concerning 43 CACS patients with a mean age of 38 years (range 14e73), the most predominant symptoms were postprandial pain (present in Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice

A

a

b Oesophagus

Oesophagus Median arcuate ligament Median arcuate ligament

Celiac Trunc Celiac Trunc

Celiac Plexus

B

Celiac Plexus

a

b

c

Figure 2 Coeliac artery compression syndrome. (A) A diagrammatic representation of the compression of the coeliac artery by the median arcuate ligament (a) or by the coeliac plexus (b). This is an adapted figure from ‘The clinical anatomy of coeliac artery compression syndrome’ by Loukas et al.19 (B) Angiography of the coeliac artery compression syndrome. Magnetic resonance angiography (a) shows a subtotal occlusion of the coeliac artery as indicated by the white arrow. In the digital substraction arteriography performed in the same patient, the compression of the coeliac artery is shown to be dependent on breathing, with a subtotal occlusion seen at expiration (b) (closed arrow) and a 60% stenosis seen at inspiration (c) (closed arrow). Also visible is the marked post-stenotic dilation of the coeliac artery (a and b) (open arrow). 80% of patients), weight loss (77%), and exerciserelated pain (40%). An abdominal bruit on physical examination was found in only 23% of patients. A history of smoking was reported by 63% of patients, other cardiovascular risk factors were present in only 6%.8 Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

In the past 50 years, various studies have shown variable results of treatment of CACS, with treatment mainly consisting of cleavage of the ligament.20e25 In small studies varying from 6 to 11 patients, the results of long-term follow-up after treatment were very disappointing, with persistent 725

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice relief of symptoms in only 0e40% of patients.20 22 25 This led many to the belief that the syndrome was non-existent, or least not due to functional coeliac artery compression, and that symptom improvement in responders was actually due to pain blockade by denervation of the vagal nerve as a result of cleavage of the ligament. Two studies providing data on surgical treatment of CACS in larger cohorts with 20 and 51 patients showed more promising results. The latter two studies showed persistent symptom relief in 68 to 75% of patients.21 23 It should be borne in mind that diagnostic options at the time of these studies were very limited and patients were often diagnosed on the basis of symptoms and an abdominal bruit, in the absence of adequate visualisation techniques and functional tests. More recently, a prospective study applied gastric exercise tonometry as the key investigation for functional assessment of mucosal perfusion in patients with suspicion of CACS. In carefully selected cases with a typical abrupt, expiration-dependent, superior compression of the root of the coeliac trunk, combined with pathological functional test result by tonometry, surgical release of the CA led to persistent relief of symptoms on long-term follow-up in 83% of patients, compared with the 0e75% in earlier studies. The existence of impaired mucosal perfusion in CACS patients was even more underlined by the fact that all successfully treated patients showed improvement or normalisation of gastric exercise tonometry after treatment.8 This was thus the ﬁrst study to show impaired mucosal perfusion in CACS, also associating symptom and functional improvement. These data are strongly supportive for the true-ischaemia hypothesis for CACS.

NON-OCCLUSIVE MESENTERIC ISCHAEMIA

Atherosclerotic narrowing of the gastrointestinal arteries is quite common. As mentioned earlier, solitary stenoses of the CA or SMA are found in 6e29% of asymptomatic subjects, rising to 69% at age >80 years. Currently, no data exist on the exact incidence of single-vessel CGI. However, one study presented data of a large cohort of CGI patients, and compared symptoms in both single- and multivessel disease patients. This showed similar symptoms in both groups, with the exception of weight loss which was signiﬁcantly more prominent in the multi-vessel disease CGI patients.6 Although no studies have so far focused speciﬁcally on outcome of endovascular treatment of single vessel atherosclerotic CGI, some conclusions can be derived from published larger, combined single- and multivessel atherosclerotic CGI cases. In patients with an isolated >70% stenosis of either CA or SMA in combination with abnormal tonometry, endovascular therapy was associated with persistent relief/ improvement of symptoms in 71 to 75% during a >1 year follow-up.6 7

The introduction of functional tests to detect mucosal ischaemia led to the recognition of another distinct group of patients.7 26 27 This group consisted of patients with symptoms compatible with CGI and with conﬁrmed mucosal ischaemia during functional testing, yet without any arterial stenosis observed with gastrointestinal arterial imaging. This condition, referred to as chronic non-occlusive mesenteric ischaemia (NOMI), is diagnosed in 13e16% of all patients with CGI.7 26 27 The exact aetiology of chronic NOMI is largely unknown, but is likely to consist of different pathophysiological mechanisms. Under physiological conditions, after a meal the intestine is able to increase the mesenteric blood ﬂow by 58e250% to meet the increased oxygen demand associated with digestion and absorption of nutrients.28 29 This effect is mediated by a decrease of mesenteric vascular resistance, which is regulated by neural, humoural and paracrine mechanisms.30 Pathological mechanisms may interfere with this regulation, making a subject unable to increase the postprandial mesenteric blood ﬂow to meet the oxygen demand, thus causing mucosal ischaemia. Mucosal ischaemia can occur in healthy subjects during strenous exercise, such as during marathon running,31 32 but this rarely causes persistent symptoms.33 In some patients, however, pathological conditions such as a chronic state of low perfusion caused by a decreased cardiac output due to heart failure, cardiac valve disease or arrhythmias, and or severe anaemia may cause chronic gastrointestinal ischaemia in the absence of vascular stenosis.34 The imbalance between oxygen supply and mucosal oxygen requirements may also be caused by other mechanisms such as spasms of small arteries,7 occlusion of smaller arteries, autonomic dysfunction, or vasculitis. The relative frequency of these different pathological mechanisms among patient with chronic NOMI is unknown, and therefore a topic of current research.

MULTI-VESSEL GASTROINTESTINAL ARTERY DISEASE AND CGI

CLINICAL PRESENTATION OF CGI Presenting symptoms

Atherosclerotic narrowing of the gastrointestinal arteries is the main cause of multi-vessel CGI. As

The clinical symptoms of CGI, whether caused by single- or multi-vessel arterial disease, and whether

Single-vessel atherosclerotic CGI

726

mentioned earlier, multi-vessel gastrointestinal arterial stenoses are found in up to 15% of asymptomatic, elderly patients.11 14 Proportionately, these stenoses lead to more than 70% occlusion of the lumen in multiple vessels. Upper abdominal symptoms in these patients should raise a high suspicion for the presence of CGI. A recent study which included gastrointestinal tonometry and a multidisciplinary approach in the work-up of all cases referred to a CGI unit showed pathological mucosal perfusion in 80 and 100% of patients with, respectively, two- and three-vessel atherosclerotic stenotic disease.6 Details on treatment options and outcome in patients with multi-vessel CGI are presented in the following paragraphs.

Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice Table 1 Clinical symptoms and presentation of patients with chronic gastrointestinal ischaemia Symptom

Total CGI N[137*

SVIDy N[84

MVID N[53

Postprandial pain Weight loss Exercise-related pain Abdominal bruit Diarrhoea Risk factor(s) for CVD Triad ‘abdominal angina’z

85% 76% 44% 25% 7% 72% 22%

84% 74% 44% 26% 4% 65% 19%

86% 78% 43% 24% 11% 82% 28%

*Patients from cited studies (Walker et al5 and Mensink et al6). yAtherosclerotic occlusive or coeliac artery compression syndrome single-vessel ischaemic disease, zCombination of postprandial pain, weight loss due to fear of eating, and an abdominal bruit. CGI, chronic gastrointestinal ischaemia; CVD, cardiovascular disease; MVID, multi-vessel ischaemic disease; SVID, single-vessel ischaemic disease.

induced by atherosclerotic or eccentric disease, are similar.6 Postprandial pain, weight loss, exerciserelated pain, and an abdominal bruit are presenting symptoms found in, respectively, 85%, 76%, 44%, and 25% of patients with CGI (see table 1).6 8 The postprandial symptoms occur typically within 15e30 min after a meal and last for 60e120 min. Recent studies have shown that only a minority of patients present with the classical ‘abdominal

angina’ triad of postprandial pain, weight loss due to fear of eating, and an abdominal bruit.6e8 Other symptoms that are occasionally found include diarrhoea, malabsorption and nausea, often in combination with postprandial pain. During upper endoscopy, oedema and erythema of the gastric mucosa can be observed in 35% and 42% of patients, respectively. However, these ﬁndings are seldom discriminative to other pathological conditions. Other endoscopic ﬁndings such as atrophy of the duodenal mucosa, and non-Helicobacter pylori/nonNSAID gastric or duodenal ulcers, are observed in only a minority of patients with CGI (ﬁgure 3).

Risk factors The risk factors for atherosclerotic gastrointestinal arterial disease seem to be compatible with the known risk factors for cardiovascular disease. The Framingham Heart Study and other large-scale epidemiological studies have identiﬁed major cardiovascular risk factors such as hypertension, dyslipidaemia, glucose intolerance, obesity, hyperhomocysteinaemia and smoking.35e37 These factors also seem to predispose to gastrointestinal arterial disease and CGI. In a cohort study of 168 patients with CGI, the risk factors for gastrointestinal arterial atherosclerosis were compatible with the

Figure 3 Endoscopic images of patients with chronic gastrointestinal ischaemia. Endoscopic images of the stomach (AeC) and duodenum (DeF) of patients analysed for chronic gastrointestinal ischaemia. In the majority of patients analysed for chronic gastrointestinal ischaemia no specific mucosal abnormalities are detected (A and D). Oedema and erythema of the gastric mucosa is observed in 35% and 42% of the patients with CGI, respectively. This was, however, shown not to be discriminative with other conditions. In 8%, a Helicobacter pylori-negative and NSAID-negative mucosal ulcer is observed in the antrum (B and C). The duodenum seldom shows pathology (D). In two patients with a total occlusion of the coeliac artery and the superior mesenteric artery ulcerations were seen in the antrum en corpus of the stomach (B and C), and whitening, and sometimes bluish colorisation, of the Kerckring folds (E and F). Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

727

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice general established risk factors for arterial atherosclerotic disease. However, the female preponderance (65%) and the high prevalence of hyperhomocysteinaemia (45%) were remarkable. These same ﬁndings were recently reported by our group, conﬁrming the female preponderance (66%) and atherosclerotic risk factors.38

DIAGNOSIS OF CGI Consensus diagnosis and multidisciplinary approach The diagnosis of CGI is currently based on the outcomes of three main components. This ﬁrst includes the medical history, symptom assessment, and physical examination of the patient. The second component consists of radiological visualisation of the gastrointestinal arteries, and the third component aims at functional testing of mucosal perfusion. Different diagnostic strategies have been suggested.6 39 40 The combination of gastrointestinal tonometry and duplex ultrasound of the abdominal arteries is currently suggested as the strategy with the highest diagnostic accuracy.39 As a referral centre, we use a standardised checklist for a patient’s symptoms and a physical examination, visualisation of the gastrointestinal arteries by means of CTeangiography, and visible-light spectroscopy for detection of mucosal hypoxaemia. The outcomes of the different diagnostic tests are, without exception, discussed in a multidisciplinary team, with an interventional radiologist, a vascular surgeon and a gastroenterologist. The team thus comes to a consensus diagnosis on the presence or absence of CGI, as well as a treatment plan if applicable. The reason for this team approach lies in the multidisciplinary character of diagnosis and treatment of this condition, with the aim to optimise management.

Visualisation of gastrointestinal arteries and blood flow Abdominal duplex ultrasound scanning, computed tomography angiography (CTA), and magnetic resonance angiography (MRA) are the current noninvasive diagnostic methods for gastrointestinal arterial stenosis. Angiography of the gastrointestinal arteries remains the ‘gold standard’ for diagnosing and staging of gastrointestinal arterial stenosis. The advantages of digital subtraction angiography (DSA) are its high sensitivity and speciﬁcity for detection of stenoses in the origin of the abdominal arteries, combined with the good quality of visualisation of the peripheral mesenteric vasculature, the possibility to assess any collateral circulation, and the ability to perform good quality imaging during inspiration and expiration. The latter is of special importance in patients suspected of CACS (ﬁgure 2B). Lateral (conventional) arteriography is the primary modality to detect extrinsic compression by the median arcuate ligament compatible with CACS. The syndrome is characterised by a sharp, superior impression of the root of the coeliac artery, which compression typically increases during expiration and decreases during 728

inspiration (ﬁgure 2B). The main disadvantages of DSA are its invasive character, the lack of visualisation of surrounding tissues, and potential complications, in particular contrast allergy and renal function impairment. Abdominal duplex ultrasound has a fairly good sensitivity and speciﬁcity for detection of stenoses in the origin of the gastrointestinal arteries. The sensitivity for signiﬁcant (>50e70%) stenoses of the CA is 75e100%, with a speciﬁcity of 88e89%. The sensitivity for signiﬁcant stenoses of the SMA is 89%, with a speciﬁcity of 92e97%.41e44 Duplex ultrasound performed during deep expiration can demonstrate a marked increase in ﬂow velocity at the compressed region of the CA, supporting the diagnosis of CACS.45 A shortcoming of abdominal duplex ultrasound as a screening tool for stenoses in the gastrointestinal arteries is the poor arterial visualisation in 10e20% of patients due to over-projection of abdominal air. Furthermore, a high sensitivity and speciﬁcity for detection of arterial stenosis is only obtained in experienced hands. In recent years CTA and MRA have evolved as promising techniques for assessment of the patency of the gastrointestinal arteries and evaluation of patients suspected of CGI. The main advantages of both imaging techniques are the non-invasive approach and the possibility to visualise the surrounding structures (ﬁgure 4). Only a few studies compared CTA or MRA with DSA as the current gold standard in the evaluation of patients suspected for CGI. In a recently published study, 52 patients had both CTA and DSA for visualisation of the gastrointestinal arteries, showing a sensitivity of CTA for abdominal arterial stenosis of 82%, with a speciﬁcity of 100%.46 Another study compared MRA and conventional DSA in 65 patients suspected for CGI. In this study the sensitivity of MRA for abdominal arterial stenosis was 100%, with a speciﬁcity of 95%.47 The MRA tended to over-rate stenoses in up to 15% of patients, and evaluation of the IMA was only possible in 64% of patients. No studies are yet available comparing CTA and MRA in patients suspected of CGI. However, it seems that the IMA and the smaller peripheral mesenteric vessels are currently better assessed with CTA than with MRA because of the higher spatial and temporal resolution and faster acquisition times of CTA.48 In addition, CTA allows the identiﬁcation of calciﬁed plaques. The main advantages of MRA over CTA are its lack of radiation exposure and the possibility to perform ﬂow measurements. MRA measurements have shown a consistent relationship between ﬂow in the portal or superior mesenteric vein and ﬂow in the arteries supplying those veins.49 The assessment of ﬂow velocities of the portal and superior mesenteric vein before and after oral caloric stimulation seems another promising diagnostic tool for CGI.

Detection of mucosal ischaemia Mucosal hypoperfusion, or hypoxaemia, is the ﬁrst sign of reduced/insufﬁcient blood ﬂow in the Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice A

a or t a

B

a o rt a

CA

CA kidney

kidney SMA kidney SMA

IMA Figure 4 Three-dimensional reconstruction of a computed tomography angiography scan. A sagittal reconstruction of a CT-angiography (panel A) showing a subtotal occlusion of the coeliac artery (CA) with post-stenotic dilation, a total occlusion at the origo of the superior mesenteric artery (SMA) over a distance of 2 cm as indicated by the arrowhead (the artery is missing over a distance of 2 cm). This has been a chronic condition, as in panel B the inferior mesenteric artery (IMA) is hypertrophied with retrograde filling of the branches of the SMA. The blue arrowheads show the meandering artery providing the large and small intestine of blood. gastrointestinal arterial tract. The blood ﬂow within the intestinal wall is unevenly distributed, and may show marked variation between mucosal and serosal sides. The mucosal side, being the metabolically most active, is relatively protected against ischaemia at disadvantage of the serosal layers.50 However, despite this preserved mucosal blood ﬂow, the mucosal villi are most vulnerable to a hypoxaemic state.51 52 This is a consequence of the microvascular situation, with counter-current oxygen exchange occurring over the length of the villus, resulting in hypoxaemia at the tip and normoxaemia at the crypts. This hypoxaemic state is associated with an anaerobic metabolic state, which eventually leads to intracellular accumulation of CO2 due to buffering mechanisms. Subsequently, this intracellular CO2 diffuses into the bowel lumen leading to an increase in intraluminal CO2. This intraluminal CO2 can be measured by means of intraluminal tonometry. For this purpose, a balloon-tipped nasojejunal cathether is placed. After inﬂation of the semi-permeable balloon, the balloon CO2 pressure equilibrates with the luminal CO2 pressure. Repeated aspiration of the gas content of the balloon during fasting and after meals thus provides an insight in the local mucosal CO2 pressures and can correlate these with symptoms.53 54 The same technique can also be applied in the colon. Currently, gastrointestinal tonometry is the only established functional diagnostic test to detect gastrointestinal mucosal ischaemia. In two large patient cohorts evaluated for possible CGI, gastrointestinal tonometry proved to have a sensitivity and speciﬁcity of, respectively, 78e85% and Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

82e92% for the detection of CGI (ﬁgure 5).6 7 Currently, gastrointestinal tonometry is used for diagnosing mucosal ischaemia by two different approaches. The ﬁrst approach introduced was gastric exercise tonometry which uses sub-maximal exercise as a provocation of gastrointestinal ischaemia, in essence very similar to the concept of exercise testing commonly used for the evaluation of cardiac ischaemia. Because gastric exercise tonometry is quite cumbersome and cannot be performed in up to 20% of patients due to their inability for physical exercise, an alternative tonometric approach was introduced. This method consisted of prolonged (24 h) combined gastric and jejunal tonometry using meals as provocation of gastrointestinal ischaemia.55 This prolonged tonometry test was found to be as accurate as gastric exercise tonometry for detection of gastrointestinal ischaemia with a sensitivity of 77% and a speciﬁcity of 94%, and is easier to perform without patient restrictions.55 Despite these considerable improvements in diagnostic tools, a diagnosis of CGI remains a challenge in clinical practice. Currently, the main obstacles for a correct and timely diagnosis of CGI are ﬁrst that symptoms are in most cases only present for a limited period of time after a provocation, and second the limited accessibility and burden of tonometry as the proposed preferred diagnostic tool. Gastrointestinal tonometry is a cumbersome and patient-unfriendly procedure to perform, and proved to be false-negative in up to 23% of CGI patients.6 7 55 Recently, an alternative diagnostic method has been introduced, so-called visible-light spectroscopy 729

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice

Figure 5 Example of 24 h tonometry in healthy subject (A) and patient with chronic gastrointestinal ischaemia (B). 24 h tonometry is performed by means of endoscopically placed nasojejunal balloon catheter. During a period of 24 h the intraluminal PCO2 is measured in both the stomach and jejunum, both in the fasting condition and with meals. The patient is asked to eat a test meal. The patient is requested to note time and duration of any symptoms in a logbook throughout the test. In panel A, the 24 h reading of the intraluminal PCO2 in the stomach (blue line) and the jejunum (purple line) of a healthy subject. In panel B, the 24 h PCO2 reading of a patient with mucosal ischaemia of the jejunum after test meals, with a close association between the pain and a rise in intraluminal PCO2. (VLS). This relatively new and promising technique allows the direct measurement of mucosal tissue oxygen saturation during endoscopy.56 Brieﬂy, a 1.5 mm probe can be introduced through the working channel of an endoscope, just above the gastric or duodenal mucosa. By analysing the backscattered light the capillary oxygen saturation can be calculated (ﬁgure 6). A ﬁrst study performed by our group showed promising results of application of VLS in patients referred for evaluation of possible CGI.27 57 In this study, cut-off values were determined in a test cohort and subsequently validated in a second cohort of patients. This showed 730

that VLS had a sensitivity and speciﬁcity of, respectively, 90% and 60% for correctly diagnosing CGI. This is similar to, or even slightly better, than gastrointestinal tonometry. VLS measurements during upper endoscopy are thus likely to become the preferred functional test method to assess patients clinically suspected for CGI, replacing gastrointestinal tonometry as functional test for mucosal perfusion. This new and easy to use VLS technique may lower the threshold for gastroenterologists in the ﬁeld to test for this disease entity. Furthermore, VLS during upper endoscopy is considerably more patient-friendly than gastrointestinal tonometry, as Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice

Figure 6 Mucosal oxygen saturation measurement of the duodenal and gastric mucosa with visible-light spectroscopy. A 1.5 mm probe (A) can be introduced through the working channel of an endoscope to be held just above the mucosa (B). The probe transmits visible white light for tissue illumination and a fibre optic bundle that directly transmits back to the monitor for continuous, real-time saturation measurements. During measurement, the endosocopy light source is switched off. The T-Stat reports capillary-weighted oxygen saturation (C). Our group recently validated mucosal oxygen saturation cut-off values for detecting mucosal ischaemia (D).57 it is less invasive, less time-consuming, and can be performed under conscious sedation on an outpatient basis (ﬁgure 6).

TREATMENT OF CGI Treatment of CGI is directed at relief of symptoms, improvement of nutritional status, and prevention of mesenteric infarction and related morbidity and mortality. A complicated course of disease (that is, bowel infarction) mainly occurs in patients with multi-vessel disease and not in those with single-vessel disease. In an American study, following 15 patients with signiﬁcant three-vessel disease for a period of 6 years, 86% progressed from asymptomatic to symptoms of abdominal pain and malnutrition, with some progressing to visceral infarction and death.58 Other studies conﬁrm that chronic symptoms precede the ﬁrst acute event of mesenteric ischaemia in 43e52% of cases.58e60 It was also shown that an ‘acute on chronic’ mesenteric event is also associated with high mortality rates, often due to the insidious presentation of the acute ischaemic event in this speciﬁc patient population.58 59 In contrast, the Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

incidence of ischaemic events and ischaemia-related mortality was not increased in a cohort of 97 asymptomatic patients with single-vessel disease when compared to an age and co-morbiditymatched control population of 456 people during 6 years of follow-up.61 Treatment of single-vessel CGI therefore primarily aims at symptom relief. Revascularisation of the gastrointestinal arteries is the main treatment of CGI associated with occlusive arterial disease. This can be achieved by open or laparoscopic surgical revascularisation, or by endovascular percutaneous transluminal angioplasty (PTA) with or without stent placement. Conservative treatment can be considered in some patients with occlusive disease who are not eligible for surgery and/or endovascular revascularisation, and in those who prefer not to undergo invasive therapy.

Surgical revascularisation Since the ﬁrst successful report on surgical treatment for CGI by Shaw and Maynard in 1958,62 many case series have been reported concerning surgical revascularisation for the treatment of CGI. 731

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice Although many controversies remain on the choice of bypass route, antegrade versus retrograde, choice of conduit, use of vein versus prosthetic graft, and whether one or more vessels should be revascularised in the presence of multi-vessel disease, surgical revascularisation is generally accepted as a successful treatment for CGI. Success rates deﬁned as relief of symptoms vary between 84% and 94%, and surgical revascularisation is associated with an exceptional long-term durability (73e93%) at 5 years, and low recurrence of symptoms (7e14%). The data from the Nationwide Inpatient Sample database showed that surgical treatment of 2128 patients with CGI was associated with a 13% mortality (range 5e26%) and 38% morbidity (range 15e61%).63 These high ﬁgures may be explained by several factors. First, the majority of CGI patients are in a debilitated condition due to signiﬁcant malnutrition and weight loss. This is, among others, caused by the fact that CGI can be a challenging diagnosis, with a diagnostic delay varying between 4 and 48 months (mean 11 months).64 Second, many patients with CGI have serious co-morbidities such as coronary artery disease, symptomatic peripheral vascular disease, renal impairment, hypertension and COPD. A combination of debilitated condition and co-morbidities are known to increase the risk of postoperative mortality and morbidity of major surgery in general.65 Third, individual centres and physicians usually have limited experience with mesenteric revascularisation, as also illustrated by most case series including not more than 40 patients in a total time span of over 10e15 years. Skills and learning curves are thus likely to have inﬂuenced the outcome of interventions. The major complications of surgical revascularisation in CGI are acute renal failure (11%), gastrointestinal infarction requiring bowel resection (8%), cardiac arrest (6%), respiratory complications (5%), myocardial infarction (5%) and haemorrhage (3%).63

Endovascular revascularisation PTA with or without stent placement of gastrointestinal arteries was reported for the ﬁrst time in 1980.66 The arteries are approached via a transfemoral or trans-brachial route, and the ostia of the gastrointestinal arteries and collateral circulation can be visualised during selective DSA of the three main gastrointestinal arteries. Total vascular occlusion was initially thought to be a contraindication for endovascular revascularisation due to plaque defragmentation and subsequent distal embolisation, but recent studies have shown its feasibility and safety in revascularisation of occluded gastrointestinal arteries67 68 (ﬁgure 7). As the length of the stenosis is an indicator of atherosclerotic plaque burden, the risk of distal embolisation is thought to be increased in patients with long segment (>2 cm) occlusion. In this case, open surgical revascularisation should be considered as ﬁrst treatment option. Another important exclusion for endovascular treatment is the presence of CACS. Due to the dynamic and non-circular nature of CACS, stents break easily in these patients, often within the ﬁrst months after placement69e73 Endovascular revascularisation is associated with a high primary technical success rate of 90e100%, with symptom relief in 73e99% of patients (tables 2 and 3). However, the relapse rate due to restenosis of the artery or in-stent stenosis is around 28% (range 11e39%), necessitating new endovascular interventions (tables 2 and 3). In earlier studies, endovascular revascularisation was reserved for patients who were not eligible for surgical revascularisation. A recent systematic review including 328 patients from 16 case series, treated with PTA for CGI, showed a PTA-associated morbidity and mortality rate of 9% and 3%, respectively. The study presented from the Nationwide Inpatient Sample database showed a morbidity rate of 20% and a mortality rate of 3.7%.63 The main complications reported after endovascular therapy are

Figure 7 Percutaneous transluminal angioplasty with stenting of a total occlusion of the coeliac artery (CA). A pigtail catheter is positioned in the aorta and a diagnostic arteriogram is performed showing total occlusion of the coeliac artery with no contrast filling of the downstream branches (A). A guidewire is passed through the stenosis (A). A balloon expandable stent (arrow) is positioned over the stenosis (B). The stent is deployed (arrowhead) (C), with filling of the downstream branches with a control arteriogram (C). 732

Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice Table 2 Outcome of treatment in patients with chronic gastrointestinal ischaemia: results of endovascular revascularisation

Study

Vessels Clinical success rate Complications Patients Total Multi Single Follow-up Direct Primary Secondary Overall Mortality N N N (months) % % % % %

Silva81 Peck82 Atkins83 Kasirajan84 Sarac67 Fioole75 Matsumoto72 Van Wanroij76 Oderich74 Dias85

59 49 31 28 65 51 33 27 83 43

56 39 26 27 65 44 NA 13 83 43

3 10 5 1 0 7 NA 14 0 0

38 37 15 24 12 25 38 19 30 43

88 92 90 NA 85 73 94 67 99 96

82 61 77 56 89 78 73 67 56 67

95 92 93 60 96 86 88 81 87 88

5 16 42 18 31 10 15 11 18 16

3 2 3 11 8 0 0 0 2 5

NA, data not available.

related to the access point, and in particular include thrombosis, dissection and haemorrhage. Complications related to the treatment of the vascular stenosis itself are infrequently reported, occurring in 2e16% of patients, the most frequent being dissection and distal embolisation due to plaque defragmentation. Systemic treatment-related complications reported are renal insufﬁciency, myocardial infarction and anaphylactic reactions. After stent placement, patients usually require anticoagulant therapy to prevent stent occlusion. At present, the majority of patients receive clopidogrel in combination with aspirin for a period of 1e3 months. There are as yet no studies in gastrointestinal arterial disease which prove that this approach prolongs stent or bypass patency, but such an effect is assumed based on data in other stenotic vascular conditions such as coronary artery disease, renal artery disease and peripheral vascular disease. Recent technical advancements, in particular the introduction of new, partially covered endovascular stents may improve stent patency even further in future.

Comparing endovascular and surgical revascularisation therapy There are currently no prospective studies that compared surgical and endovascular revascularisation in CGI patients. The incidence of the disease Table 3 Outcome of treatment in patients with chronic gastrointestinal ischaemia: results of open surgical revascularisation

Study

Clinical success rate Complications Vessels Patients Total Multi Single Follow-up Direct Primary Secondary Overall Mortality N n N % % % % (months) %

Mell86 80 Atkins83 49 Kasirajan84 85 Kruger64 39 Park87 98 Mateo88 85 McAfee89 58 Cunningham90 74 Mensink6 36 Kieny91 60 146 Oderich74

72 47 81 24 76 81 54 73 27 51 224

8 2 4 15 22 4 4 1 9 2 5

46 42 60 39 22 NA 40 71 43 102 36

NA, data not available.

Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

97 92 NA 100 98 100 96 88

86 77 87 95 94 74 90 86

NA 83 NA 97 97 81 96 95

26 39 33 13 26 45 36 43

NA 100

91 89

95 94

NA 36

4 2 8 3 5 8 10 12 28 3 3

and the diversity in surgical approaches make it unlikely that a direct comparison of both therapy regimens will be performed in the forthcoming years. Most of the knowledge and guidelines thus have to come from case series. Some studies compared cases treated with endovascular revascularisation to historical case series with surgical repair. These comparisons are, however, hampered by differences in patient selection, and improvements in surgical and endovascular techniques over time. For example, endovascular revascularisation was earlier mainly applied in patients with highrisk co-morbidities, or as a ‘bridge to surgery’ in severely debilitated patients. The available retrospective comparative studies showed similar early outcomes with endovascular and surgical revascularisation, but better long-term patency and symptom relief in the surgically treated patients. Symptoms recurred on average in 28% of the endovascular treated patients, against 7e14% in the surgically treated patients. This difference was mainly due to the recurrence of stenosis in the endovascular treated patients. This led to the general conclusion that surgical revascularisation is the ﬁrst choice treatment in younger patients and in patients without signiﬁcant co-morbidity. This approach is supported by a recent American study, showing that a choice for surgical or endovascular revascularisation could be made on the basis of a pre-operative risk assessment of mortality (box 2, table 4). In this study, the mortality rose from 0.7% in the absence of risk criteria to 3.1% in the presence of one risk factor, 12.5% with two or three risk factors, and 25% if four or ﬁve risk factors were present.74 However, in recent studies, using endovascular intervention as the primary method of revascularisation, the success rate of endovascular treatment, after re-intervention in 28% of patients, was approximately 90% at 3 years follow-up.67 75 Currently, the long-term results and complications of both treatment strategies seem similar. Against this background, the choice of ﬁrst treatment depends on patient age and condition, underlying cause of CGI, and local experience. These decisions should be made on a patient-topatient base.

Treatment of single-vessel disease CGI Recent studies have shown that single-vessel CGI is more prevalent than previously thought.6 7 The aim of intervention is to decrease the burden of symptoms instead of decreasing CGI associated mortality and morbidity, as mortality is mainly related to multi vessel CGI.61 At present, only a few studies have provided information on the success rate of intervention in single-vessel CGI patients. The success rates after therapy vary widely from 65% to 83% in larger patient groups after long-term follow-up8 75 The introduction of a functional diagnostic test for mucosal hypoxaemia, that is gastrointestinal tonometry or visible-light spectroscopy measurements, improves the selection of patients who actually have CGI and will therefore be more likely to respond to revascularisation 733

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice

Box 2 Risk factors for mortality during or after open surgical revascularisation74 Age >80 years Severe pulmonary dysfunction: < FEV1<800 ml or DLCO <50% of predicted < resting PCO2 >50 mm Hg or PO2 <60 mm Hg < home oxygen therapy Severe cardiac dysfunction: < left ventricular ejection fraction <25% < NYHA class III or IV angina pectoris < myocardial infarction <90 days < cardiac stress test positive for cardiac ischaemia < severe renal insufficiency therapy.76 The addition of this functional diagnostic tool seems especially important in patients suspected of single vessel CGI, as only 54% of the patients analysed for CGI with a single artery stenosis were shown to have CGI.6

Surgical treatment of CACS Patients with CACS require surgical treatment, as the coeliac artery has to be released from its extrinsic compression by the diaphragmatic crurae, medial arcuate ligament, and/or periaortic ganglionic tissue. Treatment consists of trans-section of the median arcuta ligament and the periganglionic tissue from the coeliac trunc all the way on to the aorta. Several investigators reported that intraoperative ultrasound may support identiﬁcation and extent of the extrinsic compression, as well as monitor the decompressive effect of the transsection during the operation.77e79 Two studies have published their results of treatment in 4424 and 438 CACS patients, respectively. The investigators performed either open surgical decompression, decompression with dilatation, or decompression with reconstruction. Sustained symptom relief was achieved in respectively 68%24 and 83%8 of patients. In the American study, sustained symptom relief was higher (76%) if CACS decompression was combined with some form of revascularisation, which was either dilatation of surgical revascularisation.24 In the Dutch study, patients

Table 4 30-day and 5-year mortality risks after open surgical revascularisation in relation to the number of risk criteria (Box 2) present at baseline Number of risk criteria

30-day mortality

5-year mortality

0 1 2e3 4e5 $6

0.7 3.1 12.7 25 50

27 31 34 47 69

eligible for CA release were identiﬁed using gastric exercise tonometry, which increased the success rate of the surgical intervention.8 Furthermore, patients that presented with classic symptoms of postprandial pain, aged between 40 and 60 years, and weight loss had a better response to surgery than those without these symptoms, reinforcing the importance of proper patient selection.24 Until recently, release of the coeliac trunk in CACS patients was performed during open laparotomy. Recently, a few groups have studied the use of laparoscopic treatment of CACS. Laparoscopic decompression surgery was shown to be as safe and effective as open surgery with a decrease in the duration of admission.78

Conservative treatment

< Gastrointestinal arterial stenoses are common, with an incidence rising with

Some CGI patients are not eligible for endovascular or surgical revascularisation therapy. This can be due to the extent of the arterial disease and/or the presence of co-morbidity. For these patients, conservative treatment can be considered. Currently, no reports have been published to support the use of a pharmacological approach to patients with CGI caused by occlusive gastrointestinal arterial disease. In theory, proton-pump inhibition and dietary modiﬁcations, that is smaller, more frequent, and low-caloric meals, both reduce the metabolic demand of the gastric and duodenal mucosa, but both therapies have not been systematically studied in humans. Animal experiments have shown that the gastrointestinal blood ﬂow is inﬂuenced by the amount and especially the caloric content of meals.28 29 80 Our own clinical experience with proton-pump inhibition treatment and/ or dietary measurements as conservative therapy in these patients varies; however, some patients respond well to this approach. Long-term conservative therapy may also be preferred for some patients with single-vessel disease. Some of these patients respond well to conservative measures, and some refuse revascularisation therapy.

< In a considerable proportion of patients, arterial stenosis is associated with

Treatment of chronic NOMI

Key messages

age.

< < < <

734

CGI. Clinical symptoms of CGI vary, the classical triad of ‘abdominal angina’ is present in only minority of CGI patients. Addition of a functional test for mucosal hypoxaemia is essential for diagnosis in single-vessel disease, and of additional benefit in multi-vessel disease. Single- and multi-vessel gastrointestinal arterial stenosis can both cause CGI and can be effectively treated. Treatment of occlusive CGI consists of surgical revascularisation or endovascular stent placement, both showing acceptable long-term results.

As described above, there is a distinct group of patients with mucosal ischaemia in the absence of macrovascular pathology. Treatment of these patients primarily aims at any identiﬁable underlying cause for the chronic hypoperfusion such as heart failure, valve disease, arrythmias, anaemia or vasculitis. There is, however, also a subgroup of patients without these underlying conditions. Two caseecohort studies have been reported in abstract form, which studied treatment of NOMI with Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice vasodilating medication. The ﬁrst study treated 44 patients with chronic NOMI with either isosorbide dinitrate, ketanserin, nicorandil or doxazosin. Although side effects were common, a success percentage deﬁned as >50% decrease of pain was achieved in 63%.26 The second cohort treated 31 patients with chronic NOMI with isosorbide dinitrate 50 mg od, followed by ketanserin 40 mg od if the nitrate therapy was unsuccessful or had to be stopped due to intolerable side effects. This study reported a success rate, deﬁned as >50% decrease of pain, in 59% of cases.27 Although both reports showed promising data for the treatment of chronic NOMI, these results have to be interpreted with caution, as both studies were not designed as placebo-controlled trials.

CONCLUSION AND PERSPECTIVES Chronic gastrointestinal ischaemia is more common then previously thought. From recent studies it emerges that the classical rules concerning ‘abdominal angina’ have to be revised. Both single and multi-vessel abdominal occlusive disease can cause CGI. A high proportion of CGI patients with identiﬁable major stenoses beneﬁts from endovascular treatment or reconstructive surgery. CGI should be considered in every patient with otherwise unexplained (upper) abdominal symptoms, and/or weight loss, and/or unexplained gastric or duodenal lesions including mucosal breaks, ulcers and atrophy. Imaging of the gastrointestinal arteries for detection of stenoses can be performed by CT- or MR-angiography or, in experienced hands, duplex ultrasound. Functional tests are becoming available and signiﬁcantly improve the detection rate of patients with mucosal ischaemia. This, in particular, pertains to the recent introduction of endoscopic visible-light spectroscopy, which enables easy detection of mucosal hypoxaemia during upper endoscopy. The increased recognition of patients with CGI can lead to earlier diagnosis and improvement of patient care.

8.

9. 10. 11. 12. 13. 14. 15.

16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

Competing interests None. Provenance and peer review Commissioned; externally peer reviewed.

28. 29.

REFERENCES 1. 2. 3. 4. 5. 6. 7.

Yasuhara H. Acute mesenteric ischemia: the challenge of gastroenterology. Surg Today 2005;35:185e95. Acosta S, Wadman M, Syk I, et al. Epidemiology and prognostic factors in acute superior mesenteric artery occlusion. J Gastrointest Surg 2010;14:628e35. Scharff JR, Longo WE, Vartanian SM, et al. Ischemic colitis: spectrum of disease and outcome. Surgery 2003;134:624e9; discussion 9e30. Higgins PD, Davis KJ, Laine L. Systematic review: the epidemiology of ischaemic colitis. Aliment Pharmacol Ther 2004;19:729e38. Walker AM, Bohn RL, Cali C, et al. Risk factors for colon ischemia. Am J Gastroenterol 2004;99:1333e7. Mensink PB, van Petersen AS, Geelkerken RH, et al. Clinical significance of splanchnic artery stenosis. Br J Surg 2006;93:1377e82. Otte JA, Geelkerken RH, Oostveen E, et al. Clinical impact of gastric exercise tonometry on diagnosis and management of

Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

30. 31. 32. 33. 34.

35.

chronic gastrointestinal ischemia. Clin Gastroenterol Hepatol 2005;3:660e6. Mensink PB, van Petersen AS, Kolkman JJ, et al. Gastric exercise tonometry: the key investigation in patients with suspected coeliac artery compression syndrome. J Vasc Surg 2006;44:277e81. Derrick JR, Pollard HS, Moore RM. The pattern of arteriosclerotic narrowing of the celiac and superior mesenteric arteries. Ann Surg 1959;149:684e9. Croft RJ, Menon GP, Marston A. Does ‘intestinal angina’ exist? A critical study of obstructed visceral arteries. Br J Surg 1981;68:316e18. Jarvinen O, Laurikka J, Sisto T, et al. Atherosclerosis of the visceral arteries. Vasa 1995;24:9e14. Bron KM, Redman HC. Splanchnic artery stenosis and occlusion. Incidence; arteriographic and clinical manifestations. Radiology 1969;92:323e8. Park CM, Chung JW, Kim HB, et al. Celiac axis stenosis: incidence and etiologies in asymptomatic individuals. Korean J Radiol 2001;2:8e13. Hansen KJ, Wilson DB, Craven TE, et al. Mesenteric artery disease in the elderly. J Vasc Surg 2004;40:45e52. Roobottom CA, Dubbins PA. Significant disease of the celiac and superior mesenteric arteries in asymptomatic patients: predictive value of Doppler sonography. AJR Am J Roentgenol 1993;161:985e8. Harjola PT. A rare obstruction of the coeliac artery. report of a case. Ann Chir Gynaecol Fenn 1963;52:547e50. Lipshutz B. A composite study of the coeliac axis artery. Ann Surg 1917;65:159e69. Dunbar JD, Molnar W, Beman FF, et al. Compression of the celiac trunk and abdominal angina. Am J Roentgenol Radium Ther Nucl Med 1965;95:731e44. Loukas M, Pinyard J, Vaid S, et al. Clinical anatomy of celiac artery compression syndrome: a review. Clin Anat 2007;20:612e17. Edwards AJ, Hamilton JD, Nichol WD, et al. Experience with coeliac axis compression syndrome. Br Med J 1970;1:342e5. Evans WE. Long-term evaluation of the celiac band syndrome. Surgery 1974;76:867e71. Mihas AA, Laws HL, Jander HP. Surgical treatment of the celiac axis compression syndrome. Am J Surg 1977;133:688e91. Watson WC, Sadikali F. Celiac axis compression: experience with 20 patients and a critical appraisal of the syndrome. Ann Intern Med 1977;86:278e84. Reilly LM, Ammar AD, Stoney RJ, et al. Late results following operative repair for celiac artery compression syndrome. J Vasc Surg 1985;2:79e91. Geelkerken RH, van Bockel JH, de Roos WK, et al. Coeliac artery compression syndrome: the effect of decompression. Br J Surg 1990;77:807e9. Bigirwamungu-Bargeman M, Geelkerken RH, Huisman AB, et al. Abdominal migraine, a new and treatable disorder mimicking functional dyspepsia. Gastroenterology 2009;136:A-773. Moons LM, Sana A, Van Noord D, et al. High diagnostic yield of direct endoscopic mucosal oxygen saturation measurements in patients suspected for chronic upper gastrointestinal ischemia. Gastrointest Endosc 2010;71:AB144. Sieber C, Beglinger C, Jaeger K, et al. Regulation of postprandial mesenteric blood flow in humans: evidence for a cholinergic nervous reflex. Gut 1991;32:361e6. Symersky T, Huisman EJ, Wasser MN, et al. Effect of fat digestion on superior mesenteric artery blood flow in humans. Clin Physiol Funct Imaging 2007;27:7e11. Matheson PJ, Wilson MA, Garrison RN. Regulation of intestinal blood flow. J Surg Res 2000;93:182e96. Van Wijck K, Lenaerts K, Van Loon LJ, et al. Functional splanchnic hypoperfusion precedes intestinal epithelial cell damage during strenuous exercise. Gastroenterology 2010;138:S-72. Otte JA, Oostveen E, Geelkerken RH, et al. Exercise induces gastric ischemia in healthy volunteers: a tonometry study. J Appl Physiol 2001;91:866e71. Kam LW, Pease WE, Thompson PD. Exercise-related mesenteric infarction. Am J Gastroenterol 1994;89:1899e900. Krack A, Richartz BM, Gastmann A, et al. Studies on intragastric PCO2 at rest and during exercise as a marker of intestinal perfusion in patients with chronic heart failure. Eur J Heart Fail 2004;6:403e7. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High

735

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice

36.

37.

38. 39. 40. 41. 42.

43. 44. 45.

46.

47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57.

58. 59. 60.

736

Blood Cholesterol In Adults (Adult Treatment Panel III). JAMA 2001;285:2486e97. Grundy SM, Brewer HB Jr, Cleeman JI, et al. Definition of metabolic syndrome: Report of the National Heart, Lung, and Blood Institute/American Heart Association conference on scientific issues related to definition. Circulation 2004;109:433e8. Smith SC Jr, Allen J, Blair SN, et al. AHA/ACC guidelines for secondary prevention for patients with coronary and other atherosclerotic vascular disease: 2006 update: endorsed by the National Heart, Lung, and Blood Institute. Circulation 2006;113:2363e72. Sana A, Van Noord D, Kooij S, et al. Risk Factors for Abdominal Arterial Atherosclerosis. Gastroenterology 2010;138:S-433. Otte JA, Geelkerken RH, Huisman AB, et al. What is the best diagnostic approach for chronic gastrointestinal ischemia? Am J Gastroenterol 2007;102:2005e10. Biolato M, Miele L, Gasbarrini G, et al. Abdominal angina. Am J Med Sci 2009;338:389e95. Moneta GL, Lee RW, Yeager RA, et al. Mesenteric duplex scanning: a blinded prospective study. J Vasc Surg 1993;17:79e84; discussion 5e6. Gentile AT, Moneta GL, Lee RW, et al. Usefulness of fasting and postprandial duplex ultrasound examinations for predicting high-grade superior mesenteric artery stenosis. Am J Surg 1995;169:476e9. Perko MJ, Just S, Schroeder TV. Importance of diastolic velocities in the detection of celiac and mesenteric artery disease by duplex ultrasound. J Vasc Surg 1997;26:288e93. Dietrich CF, Jedrzejczyk M, Ignee A. Sonographic assessment of splanchnic arteries and the bowel wall. Eur J Radiol 2007;64:202e12. Erden A, Yurdakul M, Cumhur T. Marked increase in flow velocities during deep expiration: a duplex Doppler sign of celiac artery compression syndrome. Cardiovasc Intervent Radiol 1999;22:331e2. Stueckle CA, Haegele KF, Jendreck M, et al. Multislice computed tomography angiography of the abdominal arteries: comparison between computed tomography angiography and digital subtraction angiography findings in 52 cases. Australas Radiol 2004;48:142e7. Meaney JF. Non-invasive evaluation of the visceral arteries with magnetic resonance angiography. Eur Radiol 1999;9:1267e76. Shih MC, Hagspiel KD. CTA and MRA in mesenteric ischemia: part 1, Role in diagnosis and differential diagnosis. AJR Am J Roentgenol 2007;188:452e61. Laissy JP, Trillaud H, Douek P. MR angiography: noninvasive vascular imaging of the abdomen. Abdom Imaging 2002;27:488e506. van Bommel J, Siegemund M, Henny CP, et al. Preservation of intestinal microvascular Po2 during normovolemic hemodilution in a rat model. J Lab Clin Med 2000;135:476e83. Marik PE. Sublingual capnography: a clinical validation study. Chest 2001;120:923e7. Knichwitz G, Rotker J, Mollhoff T, et al. Continuous intramucosal PCO2 measurement allows the early detection of intestinal malperfusion. Crit Care Med 1998;26:1550e7. Boley SJ, Brandt LJ, Veith FJ, et al. A new provocative test for chronic mesenteric ischemia. Am J Gastroenterol 1991;86:888e91. Groeneveld AB, Kolkman JJ. Splanchnic tonometry: a review of physiology, methodology, and clinical applications. J Crit Care 1994;9:198e210. Mensink PB, Geelkerken RH, Huisman AB, et al. Twenty-four hour tonometry in patients suspected of chronic gastrointestinal ischemia. Dig Dis Sci 2008;53:133e9. Friedland S, Benaron D, Coogan S, et al. Diagnosis of chronic mesenteric ischemia by visible light spectroscopy during endoscopy. Gastrointest Endosc 2007;65:294e300. van Noord D, Sana A, Benaron DA, et al. Endoscopic visible light spectroscopy: a new, minimally invasive technique to diagnose chronic GI ischemia. Gastrointest Endosc 2011;73:291e8. Thomas JH, Blake K, Pierce GE, et al. The clinical course of asymptomatic mesenteric arterial stenosis. J Vasc Surg 1998;27:840e4. Cho JS, Carr JA, Jacobsen G, et al. Long-term outcome after mesenteric artery reconstruction: a 37-year experience. J Vasc Surg 2002;35:453e60. Park WM, Gloviczki P, Cherry KJ Jr, et al. Contemporary management of acute mesenteric ischemia: factors associated with survival. J Vasc Surg 2002;35:445e52.

61. 62.

63. 64. 65. 66.

67. 68. 69. 70. 71.

72.

73. 74. 75.

76. 77.

78. 79.

80. 81. 82. 83.

84. 85.

Wilson DB, Mostafavi K, Craven TE, et al. Clinical course of mesenteric artery stenosis in elderly americans. Arch Intern Med 2006;166:2095e100. Shaw RS, Maynard EP 3rd. Acute and chronic thrombosis of the mesenteric arteries associated with malabsorption; a report of two cases successfully treated by thromboendarterectomy. N Engl J Med 1958;258:874e8. Schermerhorn ML, Giles KA, Hamdan AD, et al. Mesenteric revascularization: management and outcomes in the United States, 1988e2006. J Vasc Surg 2009;50:341e8 e1. Kruger AJ, Walker PJ, Foster WJ, et al. Open surgery for atherosclerotic chronic mesenteric ischemia. J Vasc Surg 2007;46:941e5. Sungurtekin H, Sungurtekin U, Balci C, et al. The influence of nutritional status on complications after major intraabdominal surgery. J Am Coll Nutr 2004;23:227e32. Furrer J, Gruntzig A, Kugelmeier J, et al. Treatment of abdominal angina with percutaneous dilatation of an arteria mesenterica superior stenosis. Preliminary communication. Cardiovasc Intervent Radiol 1980;3:43e4. Sarac TP, Altinel O, Kashyap V, et al. Endovascular treatment of stenotic and occluded visceral arteries for chronic mesenteric ischemia. J Vasc Surg 2008;47:485e91. Resch T, Lindh M, Dias N, et al. Endovascular recanalisation in occlusive mesenteric ischemiaefeasibility and early results. Eur J Vasc Endovasc Surg 2005;29:199e203. Zeller T, Rastan A, Schwarzwalder U, et al. Endovascular therapy of chronic mesenteric ischaemia. EuroIntervention 2007;2:444e51. Delis KT, Gloviczki P, Altuwaijri M, et al. Median arcuate ligament syndrome: open celiac artery reconstruction and ligament division after endovascular failure. J Vasc Surg 2007;46:799e802. Matsumoto AH, Tegtmeyer CJ, Fitzcharles EK, et al. Percutaneous transluminal angioplasty of visceral arterial stenoses: results and long-term clinical follow-up. J Vasc Interv Radiol 1995;6:165e74. Matsumoto AH, Angle JF, Spinosa DJ, et al. Percutaneous transluminal angioplasty and stenting in the treatment of chronic mesenteric ischemia: results and longterm followup. J Am Coll Surg 2002;194:S22e31. Cina CS, Safar H. Successful treatment of recurrent celiac axis compression syndrome. A case report. Panminerva Med 2002;44:69e72. Oderich GS, Bower TC, Sullivan TM, et al. Open versus endovascular revascularization for chronic mesenteric ischemia: risk-stratified outcomes. J Vasc Surg 2009;49:1472e9. e3. Fioole B, van de Rest HJ, Meijer JR, et al. Percutaneous transluminal angioplasty and stenting as first-choice treatment in patients with chronic mesenteric ischemia. J Vasc Surg 2010;51:386e91. van Wanroij JL, van Petersen AS, Huisman AB, et al. Endovascular treatment of chronic splanchnic syndrome. Eur J Vasc Endovasc Surg 2004;28:193e200. Roayaie S, Jossart G, Gitlitz D, et al. Laparoscopic release of celiac artery compression syndrome facilitated by laparoscopic ultrasound scanning to confirm restoration of flow. J Vasc Surg 2000;32:814e17. van Petersen AS, Vriens BH, Huisman AB, et al. Retroperitoneal endoscopic release in the management of celiac artery compression syndrome. J Vasc Surg 2009;50:140e7. Vaziri K, Hungness ES, Pearson EG, et al. Laparoscopic treatment of celiac artery compression syndrome: case series and review of current treatment modalities. J Gastrointest Surg 2009;13:293e8. Gallavan RH Jr, Chou CC. Possible mechanisms for the initiation and maintenance of postprandial intestinal hyperemia. Am J Physiol 1985;249:G301e8. Silva JA, White CJ, Collins TJ, et al. Endovascular therapy for chronic mesenteric ischemia. J Am Coll Cardiol 2006;47:944e50. Peck MA, Conrad MF, Kwolek CJ, et al. Intermediate-term outcomes of endovascular treatment for symptomatic chronic mesenteric ischemia. J Vasc Surg 2009;51:140e7. e1e2. Atkins MD, Kwolek CJ, LaMuraglia GM, et al. Surgical revascularization versus endovascular therapy for chronic mesenteric ischemia: a comparative experience. J Vasc Surg 2007;45:1162e71. Kasirajan K, O’Hara PJ, Gray BH, et al. Chronic mesenteric ischemia: open surgery versus percutaneous angioplasty and stenting. J Vasc Surg 2001;33:63e71. Dias NV, Acosta S, Resch T, et al. Mid-term outcome of endovascular revascularization for chronic mesenteric ischaemia. Br J Surg 2010;97:195e201.

Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

Recent advances in clinical practice 86. 87. 88.

Mell MW, Acher CW, Hoch JR, et al. Outcomes after endarterectomy for chronic mesenteric ischemia. J Vasc Surg 2008;48:1132e8. Park WM, Cherry KJ Jr, Chua HK, et al. Current results of open revascularization for chronic mesenteric ischemia: a standard for comparison. J Vasc Surg 2002;35:853e9. Mateo RB, O’Hara PJ, Hertzer NR, et al. Elective surgical treatment of symptomatic chronic mesenteric occlusive disease: early results and late outcomes. J Vasc Surg 1999;29:821e31; discussion 32.

89. 90. 91.

McAfee MK, Cherry KJ Jr, Naessens JM, et al. Influence of complete revascularization on chronic mesenteric ischemia. Am J Surg 1992;164:220e4. Cunningham CG, Reilly LM, Rapp JH, et al. Chronic visceral ischemia. Three decades of progress. Ann Surg 1991;214:276e87; discussion 87e8. Kieny R, Batellier J, Kretz JG. Aortic reimplantation of the superior mesenteric artery for atherosclerotic lesions of the visceral arteries: sixty cases. Ann Vasc Surg 1990;4:122e5.

SUBMIT

NOW

The BMJ Group is delighted to announce the launch of BMJ Open, a new and exciting open access online journal of medical research. BMJ Open publishes the full range of research articles from protocols and phase I trials to meta analyses.

Accessible to everyone

• • • •

Fully open transparent peer review Open access means maximum exposure for all articles Article-level metrics showing use and impact Rate and comment on articles

For more details visit

bmjopen.bmj.com

Gut 2011;60:722e737. doi:10.1136/gut.2009.199695

737

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

JournalScan Guruprasad P Aithal, JournalScan Editor

GI Highlights from the literature Exercise improves IBS symptoms < Johannesson E, Simren M, Strid H, et al. Physical activity improves symptoms in irritable bowel syndrome: a randomized controlled trial. Am J Gastroenterol 2011. doi:10.1038/ajg.2010.480.

Exercise appears to be of beneﬁt in some functional chronic disorders, including ﬁbromyalgia, but its effect in irritable bowel syndrome (IBS) has not been studied previously. This is the ﬁrst randomised controlled trial of exercise in IBS. Johannesson and colleagues recruited patients with IBS according to the Rome II criteria from secondary and tertiary care. Patients in the active intervention group received regular telephone advice from a physiotherapist regarding their physical activity over a 12-week period. The control arm of the trial also had supportive telephone contact with the physiotherapist, but made no lifestyle modiﬁcations. Outcomes studied included the effect on symptom severity, quality of life, anxiety and depression, and fatigue, as well as oro-anal transit time and stool form. Of the 50 receiving exercise advice, four were lost to follow-up, compared with seven of 52 in the control group. There were signiﬁcant improvements in symptom severity scores, with a 51-point reduction in mean symptom severity among those randomised to exercise versus a ﬁve-point decrease in the control group (p¼0.003) (ﬁgure 1). The proportion of patients with a reduction in symptom score of 50 or more was higher in the exercise arm, but this did not reach formal statistical signiﬁcance (43% vs 26%, p¼0.07). However, there was a signiﬁcantly higher proportion of patients assigned to control group whose symptoms worsened during the 12 weeks of the study (23% vs 8%, p<0.01). There were also signiﬁcant improvements in the physical role and functioning domains of the quality of life questionnaire among those receiving exercise advice. Differences in anxiety, depression and fatigue scores were not signiﬁcant, either within or between the groups. Interestingly, no signiﬁcant effect of exercise on oro-anal transit or stool form was demonstrated. This is the ﬁrst trial to suggest that exercise may have a beneﬁcial effect on IBS symptoms, and that physical inactivity may worsen symptom severity. The mechanism by which this is achieved is

unclear, and whether or not patients in a real-life setting will comply with advice to increase exercise levels is debatable.

Chromoendoscopydwhat is the evidence for detecting dysplasia in IBD < Subramanian V, Mannath J, Ragunath K, et al. Meta-analysis: the diagnostic yield of chromoendoscopy for detecting dysplasia in patients with colonic inflammatory bowel disease. Aliment Pharmacol Ther 2011;33:304e12.

Gastroenterologists are used to surveying colonic inﬂammatory bowel disease (IBD) but dysplasia in bowel mucosa is often multifocal and ﬂat. Dye spraying is believed to enhance visualisation of subtle mucosal abnormalities and chromoendoscopy is a newer method for detecting dysplasia in patients. A meta-analysis of published studies was performed to compare the diagnostic yield of dysplastic lesions between chromoendoscopy and standard white light endoscopy in patients with IBD undergoing surveillance colonoscopy. The research team identiﬁed six studies involving around 1277 patients. The difference in yield of dysplasia between chromoendoscopy and white light endoscopy was 7% on a per patient analysis with a number needed to treat (NNT) of 14. The difference in proportion of lesions detected by targeted biopsies was 44%, and ﬂat lesions 27% in favour of chromoendoscopy. Chromoendoscopy appears to be signiﬁcantly better than white light endoscopy in detecting dysplasia in patients with colonic IBD in the small amount of data available. But larger studies of chromoendoscopy versus standard endoscopy are clearly required.

‘Copper standard’ for Wilson’s disease < Nicastro E, Ranucci G, Vajro P, et al. Re-evaluation of the diagnostic criteria for Wilson disease in children with mild liver disease. Hepatology 2010;52:1948e56.

Early diagnosis of Wilson’s disease is rewarding because treatment can reduce signiﬁcant risk of morbidity. As children and young adults once diagnosed would be committed to lifelong treatment accurate diagnosis is crucial. Considering the limitations of individual tests, a scoring system has been proposed for the diagnosis, but this requires a liver biopsy. Nicastro et al performed a caseecontrol study including 40 asymptomatic subjects (age 1.1e20.9 years) with Wilson’s disease (conﬁrmed by an abnormal liver copper concentration plus the identiﬁcation of two disease-causing mutations or homozygosity for a single disease-causing mutation) and 58 subjects (age 1e20 years) with other liver diseases. Serum ceruloplasmin level of 20 mg/dl achieved sensitivity of 95.0% and speciﬁcity of 84.5% and a cut-off value of 40 mg/24 h of basal urinary copper excretion achieved sensitivity of 78.9% and speciﬁcity of 87.9%. The 24-h urinary copper test after penicillamine challenge performed poorly. These ﬁndings conﬁrm that serum ceruloplasmin level (<20 mg/dl) is adequate as an initial diagnostic test for Wilson’s disease in children with asymptomatic disease and 24-h urine copper excretion may support the diagnosis. However, if these are inconclusive, clinicians should proceed with a liver biopsy and employ the Wilson’s disease scoring system which has positive and negative predictive values of 93% and 91.6%, respectively.

Reviewers Alex Ford, Neeraj Bala, Guruprasad Aithal

Journals reviewed Figure 1 The IBS Severity Scoring System (IBS-SSS); subscale IBS score. 738

American Journal of Gastroenterology, Alimentary Pharmacology & Therapeutics, Hepatology Provenance and peer review Not commissioned; not externally peer reviewed. Gut 2011;60:738. doi:10.1136/gut.2011.241810

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

PostScript

LETTERS

Individualised surveillance strategies for colorectal cancer in inflammatory bowel disease We read with interest the updated guidelines for colorectal cancer (CRC) screening and surveillance in moderate and high risk groups from Cairns SR et al.1 Indeed risk stratiﬁcation is an important step forward for inﬂammatory bowel disease (IBD)-related CRC surveillance programmes. But as the authors already suggest, the adherence to these surveillance protocols is poor and, furthermore, the proposed strategy is based on risk factors originating from tertiary referral centres with high-risk patients groups. It has already been demonstrated by population-based studies that there might have been an overestimated risk of IBDrelated CRC.2 In a Dutch nested casee control study including 173 cases and 393 control patients, we identiﬁed several strong prognostic factors for IBD-related CRC in general hospitals: age, duration of primary sclerosing cholangitis (PSC) and IBD, concomitant pseudopolyps and use of anti-

Table 1

tumour necrosis factor or immunosuppressives.3 We used Poisson regression of time to CRC with time-dependent covariates in these data to estimate the individual CRC risk for patient with IBD. The different factors were weighted by their regression coefﬁcients and subsequently assigned rounded ﬁgures. For example: 3 years of PSC and 4 years of IBD both received one point in the prediction rule (table 1). Practical use of the model is illustrated with a hypothetical 50-year-old male patient, diagnosed with IBD at age 22, extensive colitis with pseudopolyps, concomitant PSC for 10 years. According to table 1, his age score is ﬁve points, and total sum is 22 points. His probability for the development of CRC in the next year is 0.2% (ﬁgure 1). Although this proposed model needs to be validated in an external cohort, we believe that is a ﬁrst, important step towards individualised surveillance for patients with IBD and can be applied in general hospitals. According to this model we would propose to start surveillance every 3 years in patients with IBD with a predicted risk of 0.2% or higher. After further validation, this model may support

Figure 1 Individualised risk of developing inflammatory bowel disease-related colorectal cancer. physicians in deciding on starting surveillance in general hospitals. Judith E Baars,1 Caspar W N Looman,2 Ewout W Steyerberg,2 Ernst J Kuipers,1,3 Christien J van der Woude1 1 Department of Gastroenterology and Hepatology, Erasmus MC, Rotterdam, The Netherlands; 2Department of Public Health, Erasmus MC, Rotterdam, The Netherlands; 3Department of Internal Medicine, Erasmus MC, Rotterdam, The Netherlands

Correspondence to Dr J E Baars, Department of Gastroenterology and Hepatology, Erasmus MC, ’s Gravendijkwal 230, Room Ba 393, 3015 CE Rotterdam, The Netherlands; [email protected] Competing interests None.

Prediction model flowchart

Step 1: Choose the number of points for each patient characteristic mentioned below: Patient’s characteristics

No. of points

Duration of IBD Duration of PSC Gender Male Female Location of IBD Leftsided colitis (UC) Extensive colitis (UC) Limited CD Extensive CD Unclassified colitis Concomitant pseudopolyps

1 point for every 4 years of IBD 1 point for every 3 years of PSC

Contributors Specific author contributions: Conception and design: CJvdW, JEB. Collection and assembly of data: JEB. Data analysis and interpretation: JEB, CWNL, EWS, CJvdW. Manuscript writing: JEB, CWNL, EWS, EJK, CJvdW. Final approval of manuscript: JEB, CWNL, EWS, CJvdW, EJK. Provenance and peer review Not commissioned; not externally peer reviewed.

2 points 0 points

Published Online First 13 December 2010 0 points 1 point 3 points 1 point 0 points 4 points

Gut 2011;60:739. doi:10.1136/gut.2010.229351

REFERENCES 1.

Step 2: Choose the number of points that matches with your patient’s age: Age

Points

Age

Points

0e6 10 37e50 6 7e9 9 51e53 7 10e12 8 54 8 13e14 7 55 9 15e17 6 56 10 18e20 5 57 11 21e23 4 58 13 24e25 3 59 14 26e27 2 60 15 28e29 1 61 16 30 0 62 18 31 1 63 19 32e33 2 64 20 34 3 65 22 35 4 36 5 Step 3: From figure 1 read your patient’s individual risk for developing inflammatory bowel disease-related colorectal cancer within the next year

Gut May 2011 Vol 60 No 5

2.

3.

Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002). Gut 2010;59:666e89. Lakatos PL, Lakatos L. Risk for colorectal cancer in ulcerative colitis: changes, causes and management strategies. World J Gastroenterol 2008;14:3937e47. Baars JE, Looman CW, Steyerberg EW, et al. The risk of inflammatory bowel disease related colorectal carcinoma is limited: results from a nationwide nested case-control study. Am J Gastroenterol 2011;106:319e28.

Efficacy of azathioprine versus mesalazine in postoperative Crohn’s diseaseeThe Authors’ response We thank Dr Ford for his comments1 on our recent paper in Gut entitled ‘Azathioprine versus mesalazine for prevention of postoperative clinical recurrence in patients with 739

Downloaded from gut.bmj.com on April 5, 2011 - Published by group.bmj.com

PostScript

Crohn’s disease with endoscopic recurrence: efﬁcacy and safety results of a randomised, double-blind, double-dummy multi-centre trial’.2 With regard to our power calculation assuming a difference in treatment effect of 35% in favour of azathioprine, although two previous studies published in 2004 did not show such an effect size,3 4 it should be noted that the patient recruitment in our study already started in February 2002 (see the Materials and methods section). Therefore, at the time our trial was designed, the results of the above-cited studies were not available for power calculation, which was based essentially on the results of the steroid-sparing trial by Candy et al.5 A posthoc power calculation would be only of minor value, as with the given results even a non-inferiority setting for azathioprine could be envisaged in retrospect. It is important to highlight that our study is the ﬁrst and, to our knowledge, the only study for prevention of postoperative clinical recurrence in patients with Crohn’s disease and proven endoscopic recurrence, and thus a population at high risk for clinical relapse. Dr Ford himself mentioned this important difference between our trial and the other studies he suggested as references for the calculations. We recommend not mixing together results from different studies with clearly different patient populations. In this context we would like to stress that the primary end point of our study was a complex one including a clinical efﬁcacy end point (Crohn’s Disease Activity Index score $200 and an increase of $60 points from baseline) and/or a safety outcome (study drug discontinuation due to an intolerable adverse drug reaction).

3.

4.

5.

and safety results of a randomised, double-blind, double-dummy multi-centre trial. Gut 2010;59:752e9. Hanauer SB, Korelitz BI, Rutgeerts P, et al. Postoperative maintenance of Crohn’s disease remission with 6-mercaptopurine, mesalamine, or placebo: a 2-year trial. Gastroenterology 2004;127:723e9. Ardizzone S, Maconi G, Sampietro GM, et al. Azathioprine and mesalamine for prevention of relapse after conservative surgery for Crohn’s disease. Gastroenterology 2004;127:730e40. Candy S, Wright J, Gerber M, et al. A controlled double blind study of azathioprine in the management of Crohn’s disease. Gut 1995;37:674e8.

Easy dye application at surveillance colonoscopy: modified use of a washing pump Pan-colonic chromo-endoscopy appears to enhance the detection of subtle colonic lesions such as ﬂat adenomas and dysplasia-associated lesions/masses (DALMs) in patients undergoing surveillance colonoscopy.1 Recent British Society of Gastroenterology guidelines now recommend pan-colonic dye spray during colonoscopy for ulcerative colitis surveillance.2 A practical drawback to the use of pancolonic dye spray is the time taken to spray each colonic segment via a catheter and then aspirate residual ﬂuid pools. Using a modiﬁed colonoscopy washing pump, we have been able to rapidly apply indigo carmine dye to the entire mucosal surface, freeing the biopsy channel to allow suction and biopsy as required. The system employs an Olympus Flushing Pump (OFP) with a foot pedal and a biopsy cap adaptor (ﬁgure 1A). Water is used during

the insertion phase to wash (and aspirate) residual faecal ﬂuid and is then exchanged for dye solution (50 ml of indigo carmine in 250 ml of water/dimethicone mixture) which can then be applied during withdrawal. Although the dye infusion does not spray onto the full circumference of the bowel wall like a diffusion catheter, full mucosal contact with the dye can be achieved easily and rapidly by air aspiration/insufﬂation and standard patient repositioning manoeuvres (ﬁgure 1B). We estimate that this technique reduces procedural time by several minutes while optimising mucosal views and biopsy access. Zacharias P Tsiamoulos, Brian P Saunders Wolfson Unit for Endoscopy, St Mark’s Hospital & Academic Institute, Imperial College, London, UK Correspondence to Dr Zacharias P Tsiamoulos, Wolfson Unit for Endoscopy, St Mark’s Hospital, London HA1 3UJ, UK; [email protected] Competing interests None. Provenance and peer review Not commissioned; not externally peer reviewed. Published Online First 15 February 2011 Gut 2011;60:740. doi:10.1136/gut.2010.228296

REFERENCES 1.

2.

Rutter MD, Saunders BP, Schofield G, et al. Pancolonic indigo carmine dye spraying for the detection of dysplasia in ulcerative colitis. Gut 2004;53:256e60. Cairns SR, Scholefield JH, Steele RJ, et al. Guidelines for colorectal cancer screening and surveillance in moderate and high risk groups (update from 2002). Gut 2010;59:666e89.

Walter Reinisch,1 Eduard F Stange,2 on behalf of the International AZT-2 Study Group 1

Abteilung Gastroenterologie und Hepatologie, Universita¨tsklinik fu¨r Innere Medizin III, Vienna, Austria; 2 Robert-Bosch Krankenhaus, Innere Medizin I, Stuttgart, Germany Correspondence to Walter Reinisch, Universita¨tsklinik fu¨r Innere Medizin III, Waehringer Guertel 18e20, Vienna A-1090, Austria; [email protected] Competing interests WR has received an unrestricted grant from Dr Falk Pharma GmbH. EFS has received speaker’s honoraria. Provenance and peer review Not commissioned; not externally peer reviewed. Published Online First 1 March 2011 Gut 2011;60:739e740. doi:10.1136/gut.2010.226928

REFERENCES 1. 2.

740

Ford AC. Efficacy of azathioprine versus mesalazine in postoperative Crohn’s disease. Gut 2010;59:1731e2. Reinisch W, Angelberger S, Petritsch W, et al. Azathioprine versus mesalazine for prevention of postoperative clinical recurrence in patients with Crohn’s disease with endoscopic recurrence: efficacy

Figure 1 (A) A modified use of Olympus Flushing Pump (OFP). (B) Sporadic adenoma in patient with ulcerative colitis highlighted by the dye. Note the background biopsies in the upper left corner.

Gut May 2011 Vol 60 No 5