COMMENTARIES 739

GORD erosive oesophagitis, this is one of the ...... initial reports to show the presence

of Barrett’s oesophagus as having a Gut: first published as 10.1136/gut.2004.052225 on 11 May 2005. Downloaded from negative impact on healing of erosive Gastro-oesophageal reflux disease: oesophagitis. Systematic biopsies were not obtained symptoms, erosions, and Barrett’s— from the oesophageal columnar seg- ment; the number of biopsies and what is the interplay? endoscopic measurement of the length of Barrett’s oesophagus were also not P Sharma standardised between participating cen- tres. Although all endoscopists were ...... trained on the LA classification system for erosive oesophagitis, the diagnosis of The presence of Barrett’s oesophagus may exert a negative impact Barrett’s oesophagus was performed on healing of erosive oesophagitis in gastro-oesophageal reflux without any predetermined criteria. disease Furthermore, obtaining biopsies from the oesophagus were left up to the discretion of the endoscopists; addi- he outcomes of patients with ero- trials of erosive oesophagitis have tional biopsies were requested but were sive oesophagitis, treated with acid excluded patients with Barrett’s oeso- not mandatory from the endoscopists. It suppression therapy (proton pump phagus and therefore the effect of T is well known that there is large inter- inhibitors), has been dictated by the healing of erosive oesophagitis in the observer variability in the endoscopic baseline severity of erosive oesophagitis, presence of Barrett’s oesophagus is not recognition of the oesophageal colum- presence of hiatus , duration of known. nar segment and that detection of therapy and, in some studies, by the In this issue of Gut, Malfertheiner intestinal metaplasia is directly related 12 7 status of the patients. and colleagues report results from to the endoscopy/biopsy technique and It has been shown that higher grades of the Progression of gastro-oesophageal number of biopsies obtained.89 More- erosive oesophagitis (Los Angeles grades reflux disease (ProGORD) trial, a large, over, it is possible that patients with C and D) have significantly lower multicentre, prospective, follow up higher grades of erosive oesophagitis healing rates as opposed to those with study of 6215 patients with reflux (grades C and D) may be more likely to lower grades of erosive oesophagitis disease treated with esomeprazole (open have been included in the ‘‘Barrett’s (grades A and B). Moreover, the major- label) (see page 746). Results for heart- group’’ as inflammatory lesions might ity of the oesophagitis trials have eval- burn resolution in patients with erosive have been mistaken as columnar areas uated healing at four and eight weeks, oesophagitis and non-erosive reflux dis- in the distal oesophagus. showing a higher proportion of patients ease (NERD) were presented for the last Complete symptom resolution, as with all grades of erosive oesopha- visit and the prognostic influence of the determined by a validated reflux disease http://gut.bmj.com/ gitis healed at week 8 compared with baseline grade of erosive oesophagitis, questionnaire, was 58.5% at two weeks 34 week 4. Similar data on healing at presence of Barrett’s oesophagus, age, and 64.8% at the last visit in the NERD .8 weeks are not consistently available sex, body mass index, and H pylori group compared with 61.1% and 70.4%, in the literature. Not only do patients infection was studied on the healing of respectively, in the oesophagitis group. with severe grades of erosive oesopha- erosive oesophagitis and, for NERD Thus the absolute difference in patients gitis have a higher degree of oesopha- patients, on complete resolution of with heartburn resolution between geal acid exposure compared with those heartburn. Barrett’s oesophagus was

the oesophagitis and NERD groups at on September 29, 2021 by guest. Protected copyright. with either no oesophagitis or low detected in 14% of patients with erosive the last visit was 5.6%, suggesting grades of oesophagitis, but they also oesophagitis and in 2.3% of NERD that these are relatively similar patient have low amplitude of oesophageal patients. The overall healing rates of groups in terms of both pathophysi- contractions and the presence of large erosive oesophagitis at eight weeks in all ology and treatment response. These hiatus .5 Therefore, it is not patients (with and without Barrett’s data however do not reflect the same surprising that the poor pathophysiol- oesophagus) was 77.5%; 79.3% in grades point in time in each group and ogy associated with severe erosive oeso- A and B compared with 69.9% in grades although the comparison is not ideal, phagitis leads to poor healing rates. C and D (p,0.0001). In patients with- this highlights the fact that complete Although a few studies have correlated out Barrett’s oesophagus, the healing symptom resolution is difficult to H pylori status with oesophagitis heal- rate of oesophagitis was 79.3% com- achieve. Symptom resolution (measured ing, with H pylori positivity associated pared with 66.7% in those with Barrett’s by validated questionnaires) can be with improved healing rates, this has (p,0.0001). These eight week healing achieved in approximately 60–75% of not been consistently documented.6 This rates in patients with Barrett’s oesopha- GORD patients treated with proton may be a phenomenon related not just gus were also directly related to baseline pump inhibitor therapy and although to the presence or absence of H pylori oesophagitis severity (78.6% in grades A the numbers may be numerically higher infection but rather to the pattern of and B; 63% in grades C and D). Heal- in patients with erosive oesophagitis, , presence of hiatus hernia, acid ing rates were lower in those with they are still nowhere closer to heal- output states, etc.2 Although patients ‘‘confirmed Barrett’s oesophagus’’ (with ing rates, suggesting that symptoms with Barrett’s oesophagus also have histological documentation of intes- are more resistant to acid suppression abnormal pathophysiology, very similar tinal metaplasia) and also those with than mucosal breaks (that is, ero- to patients with severe grades of erosive endoscopic Barrett’s oesophagus (that sions).10 On the other hand, it is not oesophagitis, the impact of the presence is, oesophageal columnar segment). clear if patients actually seek complete of Barrett’s oesophagus in patients with Whereas the presence of severe grades symptom resolution and maybe goals erosive oesophagitis has not been sys- of erosive oesophagitis (that is, C and D) such as complete resolution of symp- tematically evaluated. In fact, previous have been shown to influence healing of toms as evaluated in this and other

www.gutjnl.com 740 COMMENTARIES trials should not be the primary end repeat endoscopy may be considered 2 Sharma P, Vakil N. Helicobacter pylori and reflux disease. Aliment Pharmacol Ther point of treatment. in this subgroup of patients. Present 2003;17:297–305.

This study highlights some important drug therapy is unable to resolve 3 Castell DO, Kahrilas PJ, Richter JE, et al. Gut: first published as 10.1136/gut.2004.052225 on 11 May 2005. Downloaded from issues; firstly, symptoms, erosions, and symptoms or heal oesophagitis com- Esomeprazole (40 mg) compared with lansoprazole (30 mg) in the treatment of erosive Barrett’s can coexist in every possible pletely for this complex disease, and . Am J Gastroenterol combination in a patient with GORD, the role of other factors such as non- 2002;97:575–83. indicating that these are not indepen- acid or low acid reflux, bile reflux, 4 Richter JE, Kahrilas PJ, Sontog SJ, et al. Comparing lansoprazole and omeprazole in dent lesions; secondly, the presence of oesophageal hypersensitivity, or central onset of heartburn relief: results of a randomized, Barrett’s mucosa exerts a negative mechanisms which lead to persistent controlled trial in erosive esophagitis patients. impact on the healing of erosive oeso- symptoms, should be evaluated fur- Am J Gastroenterol 2001;96:3089–98. 5 Coenraad M, Masclee AA, Straathof JW, et al. Is phagitis; and finally, that symptom ther. Despite the major progress in our Barrettt’s characterized by more resolution is difficult to achieve in understanding of the diagnosis and pronounced acid reflux than severe esophagitis? GORD patients (with or without erosive treatment of GORD, this study high- Am J Gastroenterol 1998;93:1068–72. 6 Holtmann G, Cain C, Malfertheiner P. Gastric oesophagitis). What are the clinical lights the need for continued investi- Helicobacter pylori infection accelerates healing implications of these findings? This gation of this intriguing disease. of reflux esophagitis during treatment with the study raises questions regarding the proton pump inhibitor . Gut 2005;54:739–740. 1999;117:11–16. need for higher doses of proton pump doi: 10.1136/gut.2004.052225 7 Malfertheiner P, Lind T, Willich S, et al. inhibitors or more profound acid sup- Prognostic influence of Barrett’s oesophagus and pression in patients with Barrett’s oeso- Correspondence to: Professor P Sharma, Helicobacter pylori infection on healing of erosive gastro-oesophageal reflux disease (GORD) and phagus. Whether persistent oesophagitis University of Kansas School of Medicine, VA Medical Center, Kansas City, MO 64128, USA; symptom resolution in non-erosive GORD: report and ongoing in patients [email protected] from the ProGORD study. Gut 2005;54:746–51. with Barrett’s oesophagus can lead to a 8 Sharma P, Morales TG, Sampliner RE. Short Conflict of interest: None declared. segment Barrett’s esophagus—the need for stand- higher frequency of dysplasia and adeno- ardization of the definition and of endoscopic carcinoma remains to be evaluated and, criteria. Am J Gastroenterol 1998;93:1033–6. if this is the case, may have important REFERENCES 9 Kim R, Baggott BB, Rose S, et al. Quantitative endoscopy: precise computerized measurement chemopreventative ramifications. Sym- 1 Kahrilas PJ, Falk GW, Johnson DA, et al. of metaplas epithelial surface area in Barrett’s ptoms appear to be a poor marker for Esomeprazole improves healing and symptom esophagus. Gastroenterology 1995;108:360–6. healing of erosive oesophagitis in resolution as compared with omeprazole in reflux 10 Pace F, Bianchi Porro G. Gastroesophageal reflux oesophagitis patients: a randomized controlled disease: A typical spectrum disease (A new patients with Barrett’s oesophagus, trial. Aliment Pharmacol Ther conceptual framework is not needed). and therefore for assessing healing 2000;14:1249–58. Am J Gastroenterol 2004;99:946–9.

Gastric cancer gastritis and altered physiology which ...... the infection induces in their .6 In some patients, the infection induces http://gut.bmj.com/ an antral predominant non-atrophic Screening for early gastric cancer gastritis associated with normal or high acid secretion. Patients with this gastric K E L McColl phenotype may develop duodenal ulcers but very rarely develop gastric cancer...... Other patients develop an atrophic

pangastritis and reduced gastric acid on September 29, 2021 by guest. Protected copyright. Can serological testing for Helicobacter pylori infection and secretion and it is this gastric phenotype atrophic gastritis predict the risk of gastric cancer? which is associated with gastric cancer. These two phenotypes represent ends of a spectrum, with many subjects show- astric cancer remains one of the type and it is this which we are ing an intermediate disease pattern. It is 1 most common malignancies in discussing. also possible for patients to progress 1 Gthe world. The great majority of Major advances have been made in from the non-atrophic to atrophic pat- cases are fatal because the tumour has our understanding of the aetiology and tern with advancing years.78 usually reached an incurable stage by pathogenesis of non-cardia gastric can- In a large prospective study in Japan, 2 the time of diagnosis. The early poten- cer and of the precancerous changes Uemura et al endoscopically moni- tially curable stage of the disease is which occur in the . tored patients with H pylori infection usually asymptomatic and consequently Helicobacter pylori infection is now and different gastric phenotypes for a patients mainly present when they have accepted as an essential cofactor in the mean of 7.8 years.6 None of the duodenal developed symptoms of advanced or majority of cases of non-cardia gastric ulcer patients developed gastric cancer. complicated disease.23 There is consid- cancer.5 However, detecting the pre- However, 4.7% of non-ulcer dyspepsia erable interest in a means of detecting sence of the infection is, in itself, of patients developed cancer and its this common cancer at an early and limited value in predicting cancer occur- occurrence was strongly associated curable stage. rence; more than 50% of the world’s with the presence of atrophic pangas- In the Western world, the incidence of population have the infection and the tritis detected at the initial endoscopy. gastric cancer distal to the cardia has lifetime risk of an infected individual Such observations have stimulated fallen over recent decades whereas that developing the cancer is less than 2%. interest in using the H pylori induced at the cardia and gastro-oesophageal Recent studies have indicated that the gastric phenotype to direct screening junction has increased markedly.4 risk of a person with H pylori infection for early gastric cancer. However, at a global level, non-cardia developing non-cardia gastric cancer is H pylori infection can be diagnosed non- gastric cancer remains the predominant highly dependent on the pattern of endoscopically by means of serology.

www.gutjnl.com COMMENTARIES 741

It is also possible to detect atrophic indicate that serological screening for the cancer.16 One would therefore antici- gastritis serologically. The concentra- low pepsinogen I would allow identifi- pate that preventing people contracting

tion of pepsinogen I and the ratio of cation of a subgroup of 22% of the the infection in the first place or treating Gut: first published as 10.1136/gut.2004.052225 on 11 May 2005. Downloaded from pepsinogen I/II are low in patients population who would go on to develop the infection at an early stage before with atrophic gastritis.910 Consequently, 70% of the gastric cancers occurring irreversible atrophic gastritis develops simple blood tests can determine within the following five years. The would be the most effective and cost whether a subject has H pylori infection annual incidence of cancer in subjects effective way of reducing the incidence and whether the infection has induced in the high risk groups in this Japanese of gastric cancer. Japan and other the atrophic pattern of gastritis which based study was approximately 0.5% countries with a persistent high inci- is associated with a high risk of gastric and thus equivalent to the risk of cancer dence of non-cardia cancer would be cancer. occurrence in Barrett’s oesophagus particularly appropriate for such a pro- In this issue of Gut, Watabe and patients studied in the West.13 phylactic interventional strategy. colleagues11 present a large prospective Identifying subjects at high and However, even if every H pylori study in which they have assessed the imminent risk of developing gastric infected subject in the world had their value of combined serological testing for cancer is only useful if one can intervene infection eradicated today, the incidence H pylori infection and atrophic gastritis to improve the natural history of the of gastric cancer would probably remain in predicting the risk of gastric cancer disease. Such intervention at present high for several decades to come. This is (see page 764). The study included involves annual endoscopies to detect due to the fact that eradicating the almost 10 000 members of the general and treat the tumours at an early stage infection does not produce resolution population who had annual endoscopic of their development. In this study of atrophic gastritis and subjects with examinations for a mean of five years reported by Watabe and colleagues,11 these irreversible changes will continue following serological tests. They ana- such a strategy allowed them to detect to be at risk of gastric cancer. lysed the outcome by dividing subjects all of the cancers which developed in the Consequently, there will continue to be into four groups according to whether high risk groups while still localised a need for more efficient ways of they were H pylori seropositive or nega- within the submucosa. Approximately identifying subjects at high and immi- tive and whether they had a low 50% of the tumours were treated by nent risk of developing this sinister pepsinogen I and I/II ratio, indicating endoscopic resection and the remainder tumour. atrophic gastritis. underwent surgical operation. The abil- Gut 2005;54:740–742. The risk of developing gastric cancer ity to detect early tumours endoscopi- doi: 10.1136/gut.2004.058461 in H pylori negative patients without cally depends on adequate endoscopic atrophic gastritis was similar to that in facilities and operator skills. Although Correspondence to: Professor K E L McColl, H pylori positive patients without these are undoubtedly common place in Section of Medicine, Western Infirmary, 44 Church St, Glasgow, UK; k.e.l.mccoll@clinmed. atrophic gastritis. Relative to these two Japan, they may not be so readily gla.ac.uk groups, the risk of cancer was sixfold available in other countries with a higher in H pylori positive patients with persistent high incidence of gastric Conflict of interest: None declared. atrophic gastritis and eightfold higher in cancer. However, one great attraction

H pylori negative patients with atrophic of the pepsinogen serological screening REFERENCES http://gut.bmj.com/ gastritis. Age and male sex were also is that it allows the available resources 1 Parkn DM, Whelan SL, Ferlay J, et al. Cancer independent risk factors for cancer to be concentrated on the group most incidence in five continents, volume VIII. Lyon: occurrence. likely to benefit from them and so IARC, 2002, Scientific Publications No 155. increases cost effectiveness. 2 Allum WH, Powell DJ, McConkey CC, et al. This study provides useful confirma- Gastric cancer: a 25 year review. Br J Surg tion of the importance of the gastric Although the ability to detect gastric 1989;76:535–40. phenotype induced by H pylori infection cancer at an early curable stage is 3 Maruyama M, Kimura K. Early gastric carcinoma in determining the risk of gastric cancer. attractive, it is not always in the presenting as dyspepsia. In: Healty V, Moncur P, eds. Dyspepsia: The clinical consequences. on September 29, 2021 by guest. Protected copyright. Patients with H pylori infection but no individual’s best interests. The natural Oxford: Blackwell Science, 2000:175–86. evidence of atrophy had a similar cancer history of early gastric cancer means 4 Devesa SS, Blot WJ, Fraumeni JF. Changing risk to those with an uninfected healthy that left untreated it usually does not patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer stomach, at least over the subsequent cause clinical problems for several years 1998;83:2049–53. five years examined in this study. and indeed may not cause any symp- 5 Webb PM, Law M, Varghese C, et al. Gastric Patients with atrophic gastritis had a toms during the patient’s natural life- cancer and Helicobacter pylori: a combined 14 15 analysis of 12 case control studies nested within markedly increased cancer risk irrespec- span. Surgical gastric resection is prospective cohorts. Gut 2001;49:347–53. tive of H pylori serology. Atrophy in associated with morbidity and a small 6 Uemura N, Okamoto S, Yamamoto S, et al. those with negative H pylori serology can mortality. Although endoscopic resec- Helicobacter pylori infection and the development be largely explained by the tendency for tion is likely to be associated with less of gastric cancer. N Engl J Med 2001;345:784–9. the infection to disappear in the morbidity and mortality, the efficacy of 7 Villako K, Kekki M, Maaroos HI, et al. Chronic atrophic achlorhydric stomach.12 Due to this less radical management in achiev- gastritis: progression of inflammation and atrophy this phenomenon, one can conclude ing long term cure is less clear. in a six-year endoscopic follow-up of a random sample of 142 Estonian urban subjects. that H pylori serology on its own is of Screening and surveillance for early Scand J Gastroenterol 1991;26(suppl 186): limited value in determining the risk of gastric cancers would be more attractive 135–41. imminent development of gastric cancer if we had a simple, safe, and effective 8 Hackelsberger A, Gunthre T, Schultze V, et al. and that detecting atrophic gastritis by way of managing them. Role of aging in the expression of Helicobacter pylori gastritis in the antrum, corpus and cardia. serum pepsinogens is the important test. The ideal way to reduce the mortality Scand J Gastroenterol 1999;34:138–43. How useful would such serological due to gastric cancer is to interrupt the 9 Samloff IM, Varis K, Ihamaki T, et al. Relationship testing be in identifying the subgroup of precancerous process before an actual among serum pepsinogen I, serum pepsinogen II and gastric mucosal histology. A study in relatives the population most likely to develop cancer develops. Prevention is better with pernicious anemia. Gastroenterology cancer within the subsequent five years than cure. Current understanding is 1982;83:204–93. and who might benefit from careful that H pylori infection is usually an 10 Borch K, Axelsson CK, Halgreen H, et al. The ratio of pepsinogen A to pepsinogen C: A surveillance endoscopy? The data pre- essential cofactor in initiating the sensitivie test for atrophic gastritis. sented by Watabe and colleagues11 sequence of events that finally lead to Scand J Gastroenterol 1989;24:870–6.

www.gutjnl.com 742 COMMENTARIES

11 Watabe H, Mitsushima T, Yamaji Y, et al. antibodies in patients with advanced atrophic 15 Ubukata H, Katano M, Tabuchi T. A case of Predicting the development of gastric cancer from corpus gastritis. APMIS 2003;iii:619–24. untreated gastric cancer followed for 7 years. combining Helicobacter pylori antibodies and 13 Falk GW. Barrett’s esophagus. Gastroenterology Gastrointest Endosc 2004;60:476–80.

serum pepsinogen status: a prospective 2002;122:1569–91. 16 Brenner H, Arndt V, Stegmaier C, et al. Is Gut: first published as 10.1136/gut.2004.052225 on 11 May 2005. Downloaded from endoscopic cohort study. Gut 2005;54:764–8. 14 Tsukuma H, Oshima A, Narahara H, et al. Natural Helicobacter pylori infection a necessary 12 Kokkola A, Kosunen T, Puolakkainen P, et al. history of early gastric cancer: a non-concurrent, condition for noncardia gastric cancer? Spontaneous disappearance of Helicobacter pylori long term, follow up study. Gut 2000;47:618–21. Am J Epidemiol 2004;159:252–8.

Crohn’s disease ADIPONECTIN, AN ANTI- ...... INFLAMMATORY MEMBER OF THE C1Q/TNF SUPERFAMILY Adiponectin, a new member of the C1q/ Creeping fat in Crohn’s disease: travel- TNF molecular superfamily,15 is abun- dantly present in human sera and ling in a creeper lane of research? circulates as monomer, trimer, and high molecular weight forms. Apart from full A Scha¨ffler, H Herfarth length adiponectin, globular adiponec- tin is also biologically active.16 Recently, ...... two adiponectin receptors, hAdipoR1 Identification of a distinct secretion pattern of adipocytokines from and hAdipoR2, have been cloned.17 The signalling pathways are currently under creeping fat in Crohn’s disease and from mesenteric adipose investigation and phosphorylation of tissue in chronic inflammatory bowel diseases (IBD) or mesenteric the insulin receptor, activation of the diseases can be considered as work in progress. Characterisation AMP activated protein kinase, activation of visceral adipose tissue by its highly active secretory products of peroxisome proliferator activated receptor (PPAR)a, and modulation of may lead to the discovery of specific discrimination and activity nuclear factor kappa B (NFkB) activity markers in IBD and may provide future targets for drug therapy. In have been described as involved.18–20 addition, the cellular compartment of macrophages residing Besides its metabolic effects in the within the mesenteric adipose tissue is becoming recognised as context of hepatic insulin resistance, bearing pathophysiological relevance. type 2 diabetes mellitus, atherosclerosis, and fatty liver, it mainly exerts anti- inflammatory effects on macrophages DR BURRIL B CROHN AND THE that stores energy as triglycerides and and endothelial cells. Adiponectin can CREEPING FAT releases energy as free fatty acids. reduce secretion of TNF-a from mono-

The connective and adipose tissue However, due to the wide variety of cyte/macrophages and attenuate the http://gut.bmj.com/ 21 changes observed in patients with hormones, proteins, peptides, comple- biological effects caused by TNF-a. Crohn’s disease (CD) have received only ment factors, cytokines, enzymes, and Mice lacking adiponectin have high little attention from pathologists, receptors expressed in and secreted by levels of TNF-a mRNA in adipose 22 although fat hypertrophy, fat wrapping adipocytes, the total adipose tissue mass tissue, and viral mediated delivery of (fat creeping upon the bowel), and is currently being recognised as a real adiponectin reverses the increase in creeping fat have long been recognised endocrine organ.7–11 Thus the term adipose tissue TNF-a mRNA. In contrast 23 by surgeons as a phenomenon suitable ‘‘adipocytokines’’12 has been introduced with leptin, adiponectin prevents the on September 29, 2021 by guest. Protected copyright. for delineating the extent of active for these highly active adipocyte derived attachment of monocytes to TNF-a 24 25 disease. Dr Burril B Crohn himself, who cytokines, such as adiponectin, resistin, stimulated endothelial cells through gave his name to this chronic inflamma- leptin, interleukin 6 (IL-6), tumour downregulation of intracellular adhe- tory bowel disease, initially described the necrosis factor a (TNF-a), and many sion molecule 1, extracellular changes in the appearance of the mesen- others. Macrophages infiltrating adipose molecule 1, and E-selectin. Therefore, teric adipose tissue as a characteristic tissue can transdifferentiate from local adiponectin may inhibit the migration symptom of the disease.1 Sheehan and preadipocytes,13 suggesting the hypoth- of monocytes to the mesenteric adipose colleagues2 and others3 defined fat wrap- esis that adipocytes and macrophages tissue and suppress local TNF-a driven ping as present if more than 50% of the may be interconvertible. Charriere and proinflammatory pathways. intestinal surface is covered by adipose colleagues13 demonstrated that stroma- tissue. Fat encroachment of the anti- vascular cells from adipose tissue or THE POTENTIAL ROLE OF mesenteric surface of the bowel displays 3T3-L1 preadipocytes can transdiffer- ADIPONECTIN IN CROHN’S a characteristic feature of CD, leading to entiate to macrophages and acquire DISEASE complete enveloping of the antimesen- phagocytic activity. As these preadipo- In this issue of Gut, Yamamoto and teric surface and obliteration of the cytes express macrophage specific anti- 26 3 colleagues from the Osaka University bowel-mesentery angle. gens such as F4/80, Mac-1, CD80, CD86, School of Medicine, Japan, present an To date, the pathophysiology of creep- and CD45, preadipocytes and macro- evaluation of adiponectin secretion from ing fat has been investigated only phages may not be too different.14 The 2–5 hypertrophied mesenteric adipose tissue sporadically and it seems to have observation that adipocytes can function of patients suffering from CD (see page fallen into oblivion.6 as macrophage-like cells by expressing 789). and secreting molecules related to They demonstrated that: WHY DOES ADIPOSE TISSUE inflammation and innate immunity MATTER? directly brings the mesenteric adipose (1) tissue concentration and release of Adipose tissue has long been regarded tissue into the focus of mesenteric adiponectin (but not of IL-6) is as a passive type of connective tissue diseases. significantly elevated in CD compared

www.gutjnl.com COMMENTARIESCO 743

with patients suffering from ulcera- inflammation and thus as a chronic percentage of macrophages in adipose tive (UC) or colon cancer, feature of the disease, probably caused tissue ranges from ,10% up to .50%,44 45 suggesting a high cellular plasticity of (2) increased adiponectin secretion in by cytokine release from adjacent lym- Gut: first published as 10.1136/gut.2004.052225 on 11 May 2005. Downloaded from CD is specifically related to inflamed phoid tissues.2 Similarly, Borley and adipose tissue. MCP-1 and macrophage and hypertrophied mesenteric adi- colleagues5 interpreted the adipose tis- inflammatory protein 1a have been pose tissue (creeping fat) and not sue changes as related to the local demonstrated to be secreted with increas- to normal adipose tissue in these effects of underlying chronic inflamma- ing amounts from adipose tissue in patients, and tory infiltrates and released cytokines response to TNF-a43 44 46 and could (3) hypertrophied adipose tissue in CD such as transforming growth factor b1 therefore function as chemoattractants and TNF-a. Both cytokines have been directing macrophage precursors into becomes infiltrated by large amounts 45 47 of monocytes/macrophages. discussed in relation to stimulating stores of fat tissue. Subsequently, a proliferation and activation of mesen- permissive microenvironment created While TNF-a inhibits adipogenesis by chymal tissues.36–39 TNF-a and PPARc by adipose tissue secretion of MCSF42 downregulation of C/EBPa,PPARc,27 28 mRNA expression is increased in could lead to a continuing process of and macrophage colony stimulating fac- mesenteric adipocytes contiguous with differentiation, transdifferentiation, and tor (MCSF),29 activation of PPARc by the involved intestine in patients with maturation of preadipocytic and non- synthetic (glitazones) and endogenous CD40 compared with adipocytes contig- preadipocytic macrophage precursor ligands (15d-PG-J2)reducesTNF-a and uous with healthy intestine or with cells. As the creeping fat in CD is leptin expression and increases adiponec- controls. becoming infiltrated by a significant tinexpressioninadipocytes.30 In detail, However, specific overexpression of amount of macrophages, the cellular PPARc agonists inhibit the expression of PPARc,40 adiponectin, TNF-a,40 leptin,41 compartment of macrophages residing proinflammatory cytokines such as IL-1b, and MCSF42 in mesenteric adipocytes within the mesenteric adipose tissue is IL-2, IL-6, IL-8, monocyte chemoattrac- from patients with CD indicates that becoming recognised as bearing patho- tant protein (MCP-1), TNF-a,andmatrix adipose tissue is an effector in the physiological relevance in IBD. metalloproteases by transcriptional reg- pathogenesis of CD. Taken together, ulation and interference with signalling mesenteric adipose tissue hypertrophy ADIPONECTIN AND pathways such as NFkB (p65, p50), AP-1 can be regarded as a cause of, or as a ADIPOCYTOKINES IN (fos/jun), mitogen activated protein consequence of, intestinal inflammation GASTROENTEROLOGY kinase cascade, and STAT-1/STAT-331 32 in CD. The presence of mesenteric As adipose tissue hypertrophy is only in monocytes/macrophages, endothelial obesity at the onset of the disease, the seen in CD, secretory factors specifically cells, smooth muscle cells, and adipo- axial polarity of inflammation, the expressed in adipose tissue could possi- cytes. These data could provide the association between connective tissue bly serve as local or systemic activity potential mechanism of an anti-inflam- changes and transmural inflammation, markers for the disease or as discrimi- matory action of PPARc ligands in the and the release of highly active mol- nating markers for the diagnosis (for context of IBD and creeping fat31 and ecules from local adipocytes supports a example, differential diagnosis between could potentially be used for both redu- more active role of adipose tissue in the CD and UC). Release of highly active cing the release of proinflammatory pathogenesis of CD. proinflammatory cytokines from fat cell http://gut.bmj.com/ cytokines and increasing the release of necrosis in may explain the anti-inflammatory cytokines such as adi- VISCERAL ADIPOSE TISSUE severe disease course. In addition, the ponectin from visceral adipose tissue. In MACROPHAGES: A NEW pathophysiological role of adipocyto- thecaseofPPARc, recent data have THERAPEUTIC TARGET? kines in mesenteric panniculitis and pointed to this nuclear hormone receptor Xu and colleagues43 and Weisberg and gastrointestinal tumours (adipose tissue as a novel anti-inflammatory mediator colleagues44 reported that adipose tissue infiltration) has to be investigated. The with broad therapeutic potential in UC future potential of adiponectin and becomes infiltrated by significant on September 29, 2021 by guest. Protected copyright. and CD.33–35 amounts of macrophages (but not lym- adipocytokines in gastroenterological phocytes or granulocytes) in the context diseases is shown in table 1. ADIPOCYTOKINES IN IBD: of obesity. They also demonstrated that Gut 2005;54:742–744. SECONDARY OR CAUSATIVE? proinflammatory cytokines are pro- doi: 10.1136/gut.2004.061531 Sheehan and colleagues2 and Smedh duced mainly by adipose tissue homed and colleagues4 interpreted fat wrapping macrophages rather than by adipo- ...... solely as a consequence of transmural cytes. It has been estimated that the Authors’ affiliations A Scha¨ffler, H Herfarth, Department of Internal Medicine I, University of Regensburg, Table 1 The future potential of adiponectin and adipocytokines in Regensburg, Germany gastroenterology Correspondence to: Dr A Scha¨ffler, Department Chronic inflammatory bowel diseases of Internal Medicine I, University of Activity markers, discrimination markers (UC—CD) Regensburg, D-93042 Regensburg, Germany; Crohn’s disease [email protected] Pathophysiology of creeping fat, drug targets for transmural inflammation Mesenteric panniculitis Conflict of interest: None declared. Pathophysiology of mesenteric inflammation, targets for diagnosis and treatment Pancreatitis Pathophysiology of retroperitoneal fat necrosis, marker of prognosis, discrimination marker for REFERENCES necrotising verus oedematous pancreatitis 1 Crohn BB, Ginzburg L, Oppenheimer GD. Gastrointestinal tumours Regional , a pathologic and clinical entity. Markers of fat tissue infiltration (staging), diagnosis of peritoneal tumour spread J Am Med Assoc 1932;99:1323–9. Liver 2 Sheehan AL, Warren BF, Gear MWL, et al. Fat- Discrimination between malignant versus benign ascites, discrimination between inflammatory versus wrapping in Crohn’s disease: pathological basis non-inflammatory ascites and relevance to surgical practice. Br J Surg 1992;79:955–8. UC, ; CD, Crohn’s disease. 3 Fazio VW. The surgery of Crohn’s disease of the small bowel. In: Allan R, Keighley MRB,

www.gutjnl.com 744 COMMENTARIESCO

Alexander-Williams J, et al, eds. Inflammatory 19 Diez JJ, Iglesias P. The role of the novel adipocyte- 33 Wada K, Nakajima A, Blumberg RS. PPARc and bowel diseases. Edinburgh: Churchill Livingstone, derived hormone adiponectin in human disease. inflammatory bowel disease: a new therapeutic 1983:452–61. Eur J Endocrinol 2003;148:293–300. target for ulcerative colitis and Crohn’s disease.

4 Smedh K, Olaison G, Nystro¨m PO, et al. 20 Yamauchi T, Kamon J, Minokoshi Y, et al. Trends Mol Med 2001;7:329–31. Gut: first published as 10.1136/gut.2004.052225 on 11 May 2005. Downloaded from Intraoperative enteroscopy in Crohn’s disease. Adiponectin stimulates glucose utilization and 34 Lewis JD, Lichtenstein GR, Stein RB, et al. An Br J Surg 1993;80:897–900. fatty acid oxidation by activating AMP-activated open-label trial of the PPAR-gamma ligand 5 Borley NR, Mortensen NJ, Jewell DP, et al. The protein kinase. Nat Med 2002;8:1288–95. rosiglitazone for active ulcerative colitis. relationship between inflammatory and serosal 21 Ouchi N, Kihara S, Funahashi T, et al. Obesity, Am J Gastroenterol 2001;96:3323–8. connective tissue changes in ileal Crohn’s disease: adiponectin and vascular inflammatory disease. 35 Desreumaux P, Dubuquoy L, Nutten S, et al. evidence for a possible causative link. J Pathol Curr Opin Lipidol 2003;14:561–6. Attenuation of colon inflammation through 2000;190:196–2002. 22 Maeda N, Shimomura I, Kishida K, et al. Diet- activators of retinoid X receptor (RXR)/ 6 Colombel JF, Dubuquoy L, Sandborn WJ, et al. The induced insulin resistance in mice lacking peroxisome proliferator-activated receptor c adipose tissue as a source of proinflammatory adiponectin/ACRP30. Nat Med 2002;8:731–7. (PPARc) heterodimer: a basis for new therapeutic signals in Crohn’s disease? In: Herfarth H, 23 Curat CA, Miranville A, Sengenes C, et al. From strategies. J Exp Med 2001;193:827–38. Feagan BG, Fo¨lsch UR, et al,eds.Targets of blood monocytes to adipose tissue-resident 36 Van Deventer SJ. Tumor necrosis factor and treatment in chronic inflammatory bowel diseases. macrophages: induction of diapedesis by human Crohn’s disease. Gut 1997;40:443–8. Dordrecht: Kluwer Academic Publishers, mature adipocytes. Diabetes 2004;53:1285–92. 37 Graham MF, Bryson GR, Diegelmann RF. 2003:99–104. 24 Ouchi N, Kihara S, Arita Y, et al. Novel Transforming growth factor beta-1 selectively 7 Ahima RS, Flier JS. Adipose tissue as an endocrine modulator for endothelial adhesion molecules: augments collagen synthesis by human intestinal organ. Trends Endocrinol Metab 2000;11:327–32. adipocyte-derived plasma protein adiponectin. smooth muscle cells. Gastroenterology 8 Kershaw EE, Flier JS. Adipose tissue as an Circulation 1999;100:2473–6. 1990;99:447–53. endocrine organ. J Clin Endocrinol Metab 25 Ouchi N, Kihara S, Arita Y, et al. Adiponectin, 38 Halloran BG, Prorok GD, So BJ, et al. 2004;89:2548–56. adipocyte-derived plasma protein, inhibits Transforming growth factor-beta 1 inhibits human 9 Rajala MW, Scherer PE. Minireview: The endothelial NF-kB signaling through cAMP- arterial smooth muscle cell proliferation in a adipocyte—at the crossroads of energy dependent pathways. Circulation growth-rate-dependent manner. Am J Surg homeostasis, inflammation, and atherosclerosis. 2000;102:1296–301. 1995;170:193–7. Endocrinology 2003;144:3765–73. 26 Yamamoto K, Kiyohara T, Murayama Y, et al. 39 Murch SH, Braegger CP, Walker-Smith JA, et al. 10 Mohamed-Ali V, Pinkney JH, Coppack SW. Production of adiponectin, an anti-inflammatory Location of tumor necrosis factor alpha by Adipose tissue as an endocrine and paracrine protein, in mesenteric adipose tissue in Crohn’s immunohistochemistry in chronic inflammatory organ. Int J Obes Relat Metab Disord disease. Gut 2005;54:789–96. bowel disease Gut 1993;34:1705–9. 40 Desreumaux P, Ernst O, Geboes K, et al. 1998;22:1145–58. 27 Zhang B, Berger J, Hu E, et al. Negative Inflammatory alterations in mesenteric adipose 11 Pantanetti P, Garrapa GG, Mantero F, et al. regulation of peroxisome proliferator activated tissue in Crohn’s disease. Gastroenterolgoy Adipose tissue as an endocrine organ? A review receptor c gene expression contributes to the 1999;117:73–81. of recent data related to cardiovascular antiadipogenic effects of tumor necrosis factor-a. 41 Barbier M, Vidal H, Desreumaux P, et al. complications of endocrine dysfunctions. Clin Exp Mol Endocrinol 1996;10:1457–66. Overexpression of leptin mRNA in mesenteric Hypertens 2004;26:387–98. 28 Rosenbaum S, Greenberg A. The short- and long adipose tissue in inflammatory bowel diseaseas. 12 Shimomura I, Funahashi T, Takahashi M, et al. term effects tumor necrosis factor-alpha and Gastroenterol Clin Biol 2003;27:987–91. Enhanced expression of PAI-1 in visceral fat: BRL49653 on peroxisome proliferator-activated 42 Levine JA, Jensen MD, Eberhardt NL, et al. possible contributor to in obesity. receptor (PPAR) gamma2 gene expression and Nat Med 1996;2:800–3. other adipocyte genes. Mol Endocrinol Adipocyte macrophage colony-stimulating factor 13 Charriere G, Cousin B, Arnaud E, et al. 1998;12:1150–60. is a mediator of adipose tissue growth. J Clin Invest 1998;101:1557–64. Preadipocyte conversion to macrophage. Evidence 29 Levine JA, Jensen MD, Eberhardt NL, et al. of plasticity. JBiolChem2003;278:9850–5. Adipocyte macrophage colony-stimulating factor 43 Xu H, Banres GT, Yang Q, et al. Chronic 14 Lehrke M, Lazar MA. Inflamed about obesity. Nat is a mediator of adipose tissue growth. J Clin inflammation in fat plays a crucial role in the Med 2004;10:126–7. Invest 1998;101:1557–64. development of obesity-related insulin resistance. 15 Shapiro L, Scherer PE. The crystal structure of a 30 Okuno A, Tamemoto H, Tobe K, et al. J Clin Invest 2003;112:1821–30. complement-1q family protein suggests an Troglitazone increases the number of small 44 Weisberg SP, McCann D, Desai M, et al. Obesity is associated with macrophage accumulation in evolutionary link to tumor necrosis factor. Curr adipocytes without the change in white adipose http://gut.bmj.com/ Biol 1998;8:335–8. tissue mass in obese Zucker rats. J Clin Invest adipose tissue. J Clin Invest 2003;112:1796–808. 16 Fruebis J, Tsao TS, Javorschi S, et al. Proteolytic 1998;101:1354–61. 45 Curat CA, Miranville A, Sengenes C, et al. From cleavage product of 30-kDa adipocyte 31 Chinetti G, Fruchart JC, Staels B. Peroxisome blood monocytes to adipose tissue-resident complement-related protein increases fatty acid proliferator-activated receptors (PPARs): Nuclear macrophages: induction of diapedesis by human oxidation in muscle and causes weight loss in mice. receptors at the crossroads between lipid mature adipocytes. Diabetes 2004;53:1285–92. Proc Natl Acad Sci U S A 2001;13:2005–10. metabolism and inflammation. Inflamm Res 46 Sartipy P, Loskutoff DJ. Monocyte chemoattractant 17 Yamauchi T, Kamon J, Ito Y, et al. Cloning of 2000;49:497–505. protein-1 in obesity and insulin resistance. Proc adiponectin receptors that mediate antidiabetic 32 Dubuquoy L, Dharancy S, Nutten S, et al. Role of Natl Acad Sci U S A 2003;100:7265–70. metabolic effects. Nature 2003;423:762–9. peroxisome proliferator-activated receptor c and 47 Robker RL, Collins RG, Beaudet AL, et al. 18 Chandran M, Phillips SA, Ciaraldi T, et al. retinoid X receptor heterodimer in Leukocyte migration in adipose tissue of mice null on September 29, 2021 by guest. Protected copyright. Adiponectin: more than just another fat cell hepatogastroenterological diseases. Lancet for ICAM and Mac-1 adhesion receptors. Obes hormone? Diabetes Care 2003;26:2442–50. 2002;360:1410–18. Res 2004;12:936–40.

Chronic pancreatitis endocrine insufficiency and a severe ...... chronic pain syndrome. Although eluci- dation of the mechanisms underlying are incomplete, Are we studying the correct state of the considerable progress has been made in our understanding of the fibrosis stellate cell to elucidate mechanisms of process as a result of identification and characterisation of pancreatic stellate chronic pancreatitis? cells (PSCs) starting in 1997.1–3 Studies with these cells suggest that they play a S J Pandol key role in chronic pancreatitis in a ...... manner analogous to hepatic stellate cells and hepatic fibrosis45 Important insights into the states of pancreatic stellate cells and In common with liver fibrosis and their relation to the disease conditions of the pancreas hepatic stellate cells there is increasing evidence demonstrating a central role for he processes of chronic pancreatitis cells of the exocrine and endocrine PSCs in pancreatic fibrosis and chronic include chronic inflammation and pancreas. These processes lead to irre- pancreatitis. In the normal pancreas, Tfibrosis with loss of parenchymal versible and debilitating exocrine and quiescent PSCs are identified using

www.gutjnl.com COMMENTARIES 745 antibodies to desmin, a cytoskeletal disease? These questions have not been to apoptosis. Furthermore, it is very protein and specific PSC marker.1 They adequately addressed. likely, as proposed by Manapov et al,that 11 Cip1/WAF1 are present in the periacinar space with The paper by Manapov and colleagues p21 plays a central role in pro- Gut: first published as 10.1136/gut.2004.052225 on 11 May 2005. Downloaded from long cytoplasmic processes encircling the in this issue of Gut provides important liferation, apoptosis, and differentiation base of the acinus. Similar to hepatic insights into the states of PSCs and their of the stellate cell. The mechanisms stellate cells in their quiescent state, PSCs relation to the disease conditions of the underlying regulation of p21Cip1/WAF1 store significant amounts of vitamin A as pancreas (see page 814). The key findings expression and translocation are un- lipid droplets in their cytoplasm. in this study are that stellate cells under- known. Establishing the underlying There is general acceptance that during going activation in culture can either mechanism for one or both of these pancreatic injury PSCs are activated in a die by apoptosis or convert to ‘‘fibro- processes may lead to potential strategies manner similar to hepatic stellate cells. blasts’’ that are resistant to apoptosis. for the treatment of fibrosing diseases. Activation consists of processes of trans- Furthermore, the authors found that the Gut 2005;54:744–745. formation to a myofibroblastic phenotype phenotype of activated stellate cells sus- doi: 10.1136/gut.2004.054080 with loss of vitamin A stores and expres- ceptible to apoptosis occurs in an experi- sion of the cytoskeletal protein a smooth mental model of pancreatitis that Correspondence to: Dr S Pandol, UCLA Department of Medicine, West Los Angeles muscle actin (a-SMA); production and resolves with return of the pancreas to VAGLAHS, 11301 Wilshire Blvd., Building secretion of large amounts of extracellular normal after removal of the injury. On 258, Room 340, Los Angeles, CA 90073, USA; matrix proteins, including collagen, fibro- the other hand, cells with the ‘‘fibroblast’’ [email protected] nectin, and laminin.6 Activation can be phenotype are present in an experimental Conflict of interest: None declared. mediated by cytokines such as transform- model of progressive pancreatic fibrosis. ing growth factor b and platelet derived The authors demonstrated that in the growth factor.2 These agents can be activated stellate cell susceptible to apop- REFERENCES produced and secreted by pancreatic tosis, a cell cycle inhibitory protein 1 Apte MV, Haber PS, Applegate TL, et al. Periacinar parenchymal cells, inflammatory cells, p21Cip1/WAF1 is present in the nucleus. stellate shaped cells in rat pancreas: identification, 78 isolation, and culture. Gut 1998;43:128–33. and PSCs themselves. These effects of With conversion of the stellate cell to the 2 Bachem MG, Schneider E, Gross H, et al. the stellate cells to produce growth factors ‘‘fibroblastic’’ state, p21Cip1/WAF1 trans- Identification, culture, and characterization of and inflammatory mediators can be locates to the cytoplasm. Cytoplasmic pancreatic stellate cells in rats and humans. Cip1/WAF1 Gastroenterology 1998;115:421–32. responsible for autocrine mediated pro- p21 binds to and inhibits act- 3 Saotome T, Inoue H, Fujimiya M, et al. Morphologi- liferation and activation of stellate cells as ivities of Rho kinase 2 and apoptosis cal and immunocytochemical identification of well as for the chronic inflammatory signal regulating kinase 1, resulting in periacinar fibroblast-like cells derived from human pancreatic acini. Pancreas 1997;14:373–82. response in chronic pancreatitis. These decreased proliferation signals and apop- 4 Li D, Friedman SL. Liver fibrogenesis and the role responses of the stellate cell may account tosis resistance. of hepatic stellate cells: new insights and for the continued progression of chronic Of interest, a-SMA, a marker of prospects for therapy. J Gastroenterol Hepatol 1999;14:618–33. pancreatitis processes in individuals, even activation in stellate cells, is reduced in 5 Friedman SL. Seminars in medicine of the Beth after cessation of alcohol abuse. expression in ‘‘fibroblastic’’ cells. Thus Israel Hospital, Boston. The cellular basis of Support for PSCs in the pathogenesis although a-SMA is a frequently used hepatic fibrosis. Mechanisms and treatment strategies. N Engl J Med 1993;328:1828–35. http://gut.bmj.com/ of pancreatic fibrosis in chronic pan- measure of the activated state of stellate 6 Mews P, Phillips P, Fahmy R, et al. Pancreatic creatitis comes from investigations of cells, it may not identify the cell stellate cells respond to inflammatory cytokines: pancreatic tissue in patients with population responsible for fibrosis. potential role in chronic pancreatitis. Gut 2002;50:535–41. chronic pancreatitis and from animal Findings that non-a-SMA expressing 7 Masamune A, Kikuta K, Satoh M, et al. Alcohol experimental models.910 The results fibroblastic cells are responsible for both activates activator protein-1 and mitogen- demonstrated that in both situations pancreatic and hepatic fibrosis have activated protein kinases in rat pancreatic stellate cells. J Pharmacol Exp Ther 2002;302:36–42. PSCs, as identified by a-SMA actin been reported in experimental models 8 Masamune A, Kikuta K, Suzuki N, et al. A c-Jun 12 13

antibodies, are present in fibrotic areas, of pancreatitis and human cirrhosis. N-terminal kinase inhibitor SP600125 blocks on September 29, 2021 by guest. Protected copyright. as determined by Sirius red or collagen I What is the significance of the find- activation of pancreatic stellate cells. J Pharmacol Cip1/WAF1 Exp Ther 2004;310:520–7. antibody staining. Furthermore, a-SMA ings related to p21 ? This cell 9 Haber PS, Keogh GW, Apte MV, et al. Activation positive cells also stained by in situ cycle inhibitory protein may represent a of pancreatic stellate cells in human and hybridisation with a probe for collagen central regulator of terminal differentia- experimental pancreatic fibrosis. Am J Pathol 1999;155:1087–95. a1 mRNA, indicating that these cells are tion, as shown in monocytic and neural 10 Casini A, Galli A, Pignalosa P, et al. Collagen 14 15 Cip1/WAF1 an important source of collagen in cells. In these cases, p21 type I synthesized by pancreatic periacinar fibrotic areas. Such findings provide expression with localisation in the stellate cells (PSC) co-localizes with lipid strong evidence for PSCs in the mechan- nucleus leads to cell cycle arrest fol- peroxidation-derived aldehydes in chronic alcoholic pancreatitis. J Pathol 2000;192:81–9. ism of pancreatic fibrosis. lowed by its translocation to the cyto- 11 Manapov F, Muller P, Rychly J. Translocation of Because of the key role for the stellate plasm where it is associated with p21Cip1/WAF1 from the nucleus to the cytoplasm cell in the mechanism of these chronic terminal differentiation. Translocation correlates with pancreatic myofibroblast to Cip1/WAF1 fibroblast cell conversion. Gut 2005;54:814–22. inflammatory/fibrosing states, elucida- of p21 to the cytoplasm in these 12 Tashiro M, Nakamura H, Taguchi M, et al. Oleic tion of the mechanisms of its transfor- cells also results in apoptosis resistance, acid-induced pancreatitis alters expression of mation is necessary in order to develop as occurred in the stellate cells. transforming growth factor-beta1 and extracellular matrix components in rats. Pancreas treatment strategies. One of the major Considering the similarities in the 2003;26:197–204. dilemmas in the research of stellate cells pattern of changes in p21Cip1/WAF1 asso- 13 Levy MT, McCaughan GW, Marinos G, et al. is the fact that they undergo transfor- ciated with differentiation in monocytic Intrahepatic expression of the hepatic stellate cell marker fibroblast activation protein correlates mation to an activated phenotype after cells, neuronal cells, and stellate cells, the with the degree of fibrosis in C virus isolation during in vitro culture. Do results in the report by Manapov and infection. Liver 2002;22:93–101. cultured stellate cells undergoing trans- colleagues11 provides a basis to propose 14 Asada M, Yamada T, Ichijo H, et al. Apoptosis inhibitory activity of cytoplasmic p21(Cip1/ formation in vitro represent the state of that the ‘‘fibroblastic’’ phenotype of the WAF1) in monocytic differentiation. EMBO J the stellate cell in chronic fibrosing stellate cell is a terminally differentiated 1999;18:1223–34. diseases of the pancreas? Is this the state. As described in this report and 15 Tanaka H, Yamashita T, Asada M, et al. 13 14 Cytoplasmic p21(Cip1/WAF1) regulates neurite correct state of the PSC to study for previous ones, this phenotype ex- remodeling by inhibiting Rho-kinase activity. J Cell determining its role in chronic fibrosing presses little or no a-SMA and is resistant Biol 2002;158:321–9.

www.gutjnl.com