Gastroesophageal reflux disease is a progressive disease

Transcription

1 Digestive and Liver Disease 39 (2007) Viewpoint Gastroesophageal reflux disease is a progressive disease F. Pace a,, S. Pallotta a, N. Vakil b,c a L. Sacco University Hospital, Milan, Italy b University of Wisconsin School of Medicine and Public Health, Madison, WI, United States c Marquette University College of Health Sciences, Milwaukee, WI, United States Received 28 March 2006; accepted 23 November 2006 Available online 23 March 2007 Abstract It is controversial whether gastrooesophageal reflux disease represents a spectrum disease from a nonerosive to a complicated one, or whether it is a categorial disease, i.e. it can be divided into three categories, such as nonerosive gastrooesophageal reflux disease, erosive gastrooesophageal reflux disease and Barrett s esophagus (BE) with little or no transition from one category to the other. This controversy might be of general interest, because it has some implications in the management of the patient. However, literature data concerning the natural history of gastrooesophageal reflux disease are very limited, and in particular very few papers have dealt with the issue of describing the natural history of patients with nonerosive gastrooesophageal reflux disease. Aim of the present review is to reassess these scanty data, and to try to demonstrate that progression from milder to more severe forms of gastrooesophageal reflux disease is possible and documented Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. Keywords: Erosive oesophagitis; GERD natural history; NERD 1. Introduction Gastroesophageal reflux (GER) is a multifaceted disease presenting with both oesophageal and extra-oesophageal clinical manifestations. It has a major impact on the workload of both the family physician and the gastroenterologist [1] due to the high prevalence in the general population. Despite the increased awareness among patients, doctors and medical authorities about the importance of this pathological condition, there are still a number of unsolved problems in its clinical management [2]. For example, many authors believe that the disease might be modelled as a spectrum one, encompassing a broad range of all clinical manifestations due to the retrograde flow of gastric contents into the esophagus [3 7]. On the contrary, another school of thought [8 11] believes that GER disease might be divided into three categories, i.e. nonerosive GERD (or NERD), erosive GERD (or ERD) and Corresponding author at: Department of Gastroenterology, L. Sacco University Hospital, Via G. B. Grassi 74, Milan, Italy. Tel.: ; fax: address: cn.fapac@tin.it (F. Pace). Barrett s esophagus with little or no transition from one category to the other. Finally, some others simply state that the evidence for the crossover (or its lack) between subcategories is simply inconclusive [12]. Resolving this controversy could be of interest for several reasons. For example, it might be useful to know if a patient without oesophageal mucosal damage (NERD) may develop Barrett s esophagus at a similar rate as patients with ERD and thus determine if endoscopic surveillance is needed for such a patient. As a consequence, the generally accepted recommendation that an upper endoscopy once in a lifetime in a patient with GERD symptoms is enough (to exclude BE), could be wrong or at least not completely supported by evidence. Additional reasons of interest are that the research on risk factors, pathophysiological mechanisms and therapeutic modalities could focus on unified strategy rather than considering each of the phenotypic presentation of GERD to be evaluated individually [13]. While the issue of the role of endoscopy and, in particular, of endoscopic surveillance in GERD is still debated [14,15], we must acknowledge the limitation of available data concerning the natural history of GERD. The aim of the present /$ Editrice Gastroenterologica Italiana S.r.l. Published by Elsevier Ltd. All rights reserved. doi: /j.dld

2 410 F. Pace et al. / Digestive and Liver Disease 39 (2007) review is to reassess available data, to show that progression from milder to more severe forms of GERD is both possible and documented. For this reason, we believe, any model based on a classification of the disease into three separate subgroups that are mutually exclusive is misleading. We will examine first the subgroup of NERD patients and then patients with erosive oesophagitis. Finally, we will propose our model of GER disease Do NERD patients progress to oesophagitis? We reported one of the first natural history studies of symptomatic GERD patients without endoscopic oesophagitis but with a pathological oesophageal ph-metry [16]. In that study, we showed that 5 of 33 such patients treated with antacids or prokinetic agents developed endoscopic oesophagitis within 6 months, and that the extent of oesophageal acid exposure at entry was not predictive of this complication. In a subsequent study [17], we extended the observation of the original patient group up to a median duration of 10 years. The first interesting observation regarding this patient sample is that almost all patients that we were able to trace (28/29) are affected by GERD symptoms when antisecretory drugs are discontinued, and therefore the majority (75%) were on such therapy due to GERD symptoms. Secondly, a very high proportion (89%) of our patients in whom repeat endoscopy was performed (N = 18) showed an erosive oesophagitis. Thus, a considerable proportion of the original patient cohort indeed showed a progression from nonerosive to erosive disease. Schindlbeck et al. [18], in a study investigating the fate of GERD patients with and without oesophagitis, reported on 16 patients with ph-documented GERD and no oesophagitis 3 years after the diagnosis. During this period, four patients (25%) developed reflux oesophagitis, while the majority of the patient population, which also included patients with oesophagitis at entry, was still taking medications on a daily basis because of their GERD symptoms. Symptoms were rated to be equal or worse than at entry by 70% of patients, in the absence of treatment. In a Finnish study, 57 consecutive referrals with symptoms of GERD were treated by modification of lifestyle/antacids [19]. Initial assessment included endoscopy and oesophageal ph recording, and patients were then followed up for a median of 19.5 years. Of the 30 patients with no evidence of erosive oesophagitis at presentation, 5 (17%) developed grade 1 oesophagitis according to Savary-Miller classification. In the study by Mc Dougall et al. [20], 71% of the 17 patients with a ph-metry documented NERD complained of frequent heartburn years after initial diagnosis, 59% were on daily acid suppressive therapy and 24% of those patients who had repeat endoscopy developed oesophagitis. Again, a progression from nonerosive to erosive GERD was observed, at least in a proportion of patients. More recently, we have performed a study on patients with typical GERD symptoms presenting in our laboratory to undergo 24-h oesophageal ph monitoring. We have analysed patients (N = 35) with a pathological investigation, defined as a 24-h percentage of GER exceeding 5.0% of the total recording time, and with a negative upper GI (gastrointestinal) endoscopy. These NERD patients have been interviewed by means of a structured questionnaire on average 3 years after the initial diagnosis in order to assess the presence and severity of GERD symptoms, the therapy (if any) received during this period of follow-up and the results of any subsequent endoscopic examination performed. The results of this retrospective survey show that 14% of those NERD patients who underwent repeat endoscopy developed erosive oesophagitis during the 3-year follow-up, despite the fact that almost all of them received effective symptomatic treatment, i.e. H2-receptor antagonists (H2- RA) or PPI therapy [21]. Finally, in a recent multicenter trial [22] conducted on 588 patients with NERD, and assessing the effectiveness of continuous versus on demand PPI maintenance therapy, it was observed that a proportion as high as 5% of patients treated on demand developed erosive changes within 6 months of study, as compared with 0% in the continuous treatment arm. A study has been conducted in a cohort of 3894 patients with predominant heartburn, with or without oesophagitis (1717 NERD, 1512 Los Angeles grade A/B and 278 LA grade C/D, and 387 had Barrett s esophagus) under routine clinical care in Germany, Austria and Switzerland (ProGERD study) [23]. After initial treatment with esomeprazole, they were followed up for 2 years, regardless of their response. Medical therapy or endoscopy was initiated at the discretion of their primary care physician, in line with routine care. At 2 years, endoscopy with biopsy was performed according to the protocol. The results were as follows: 25% of patients who had NERD at baseline progressed to LA A/B and 0.6% to LA C/D. At 2 years, 22% of patients had been off medication for at least 3 months. The conclusions of the authors were that GERD does not seem to be a categorial disease. Progression and regression (the latter likely due to therapy) between grades was observed in this large cohort of patients under routine clinical care. Another recent study has examined the possible progression in 47 subjects with symptomatic GERD without endoscopic evidence of oesophagitis, out of a group of 497 patients undergoing upper GI endoscopy for a variety of reasons [24]; all those patients ( ) were endoscopically assessed annually for 5 years. Oesophagitis developed in 36.2% of patients with NERD, as compared with 11.3% in the control group, with an hazard ratio of developing oesophagitis in the former group of The authors concluded that the condition of symptomatic GERD carries a high risk of developing oesophagitis, which increases steadily with time and was more frequent in those NERD patients with hiatus hernia who smoke and drink alcohol and who are without Helicobacter pylori (H. pyroli) infection [24]. All these studies indicate that some patients with NERD may indeed develop erosive disease at an approximate rate of about 10% per year. If this rate remains stable with time,

3 F. Pace et al. / Digestive and Liver Disease 39 (2007) a substantial proportion of patients with NERD may develop ERD within 10 years, which is a figure close to what we observed in our 10-year follow-up study of NERD patients [17]. These conclusions are in keeping with the results of a recently published systematic review which examined 22 papers on the endoscopic assessment of GERD, erosive or nonerosive, over time periods of more than 12 months [25]. In this review, authors conclude that the observed progression rate from NERD to ERD ranges in the literature from 0% to 30%. The variability may be related to the duration of follow-up and other factors such as H. pylori infection Do patients with mild oesophagitis progress to more severe grades? We will restrict ourselves to observational studies or to studies in which patients have been treated with placebo or antacids/lifestyle advice to avoid interference with more active forms of treatment. Thus, we will mainly review the literature before H2-RA or proton pump inhibitors (PPI). In the Finnish study already mentioned [19], only two of the five patients with grade 1 oesophagitis at initial presentation progressed, both to grade 2, whereas none of the patients with grade 2 progressed to more severe degrees. Patients were managed without acid-suppression therapy. In the study by Schindlbeck et al. [18], one of the eight patients with oesophagitis at entry showed a progression from grade 2 by Savary-Miller to grade 3, following a mean period of 41 months, whereas in two cases there was a regression from grade 3 to grade 1. In this study, however, only a quarter of the total sample was treated with antacids, whereas the remaining took H2-RA or even PPI. An interesting study from Japan investigated the natural history of low-grade reflux oesophagitis, i.e. grade A or B according to Los Angeles classification [26]: 105 such patients were followed up, without medical treatment, by endoscopy for a mean of 5.5 years after initial diagnosis. Following this period, 11 patients (10.5%) progressed to more severe grades, whereas 60% relapsed with a grade similar to the initial one and 29.5% had no further episodes of reflux oesophagitis. In the above quoted multinational European study called ProGERD, 1.6% of patients who had LA A/B progressed to LA C/D and 61% regressed to NERD; 42% of patients who had LA C/D regressed to LA A/B and 50% regressed to NERD (all figures exclude patients with confirmed BE at baseline). As above recalled, however, these patients were under routine therapy, which included PPI use [23]. Finally, in a retrospective study from Switzerland [27] on 701 GERD patients, who have been medically treated between 1963 and 1991 with a variety of medical therapies (including H2-RA and PPI), 23% had a relapse of greater grade than the initial diagnosis, whereas in 31% the relapse was to a less severe grade and in 46% no relapse occurred. In the paper, the duration of treatment is not stated nor the length of follow-up specified. In conclusion, the true (i.e. without therapy) natural history of reflux oesophagitis is largely unknown; the very few available studies conducted under these conditions do show some progression in the severity of oesophagitis, which is also described after medical therapy, including potent inhibition of gastric acid secretion. The authors of a systematic review give an annual progression rate ranging from 1% to 22% [25] Do NERD or ERD patients progress to BE or adenocarcinoma? In the above reported study by McDougall et al. [20], 11% of patients with erosive oesophagitis and without BE at entry showed BE at follow-up repeat endoscopy. Interestingly, Cameron et al. [28] have demonstrated that in a series of 101 patients with BE undergoing endoscopy on two or more occasions at least 1 year apart, 22 had oesophagitis on every endoscopic examination and 32 had no oesophagitis on any examination, whereas the remaining had oesophagitis at only one examination. These observations suggest that a progression from NERD or ERD to BE is possible. To our knowledge, there is only one study describing the progression from NERD to BE [29]: in this study, conducted in Germany, 2 of 34 patients with typical GERD symptoms and a normal endoscopy developed a histologically confirmed Barrett s esophagus after a mean period of 35 months. It is not clear whether patients took any medication during the follow-up period. The progression from ERD to BE has been reported in five studies on medical or surgical treatment of ERD patients, which have been recently reviewed by Fullard et al. [25]. To summarize, there is evidence that a proportion between 1% and 13% of patients with erosive oesophagitis developed Barrett s esophagus annually. It is also possible that NERD or EE (erosive esophagitis) patients might develop adenocarcinoma even in the absence of Barrett s esophagus. For example, Lagergren et al. [30] have demonstrated that the risk of oesophageal adenocarcinoma is much more related to the duration and severity of gastrooesophageal reflux disease than to the presence of BE, and therefore the latter could be considered to be a common, but not necessary step, in the evolution of oesophageal adenocarcinoma [30]. Similar observations have subsequently been made by Chow et al. [31] and Farrow et al. [32]. Conio et al. [33], in a retrospective survey conducted in the Olmsted County, were able to demonstrate that many cases of adenocarcinoma of the esophagus occurred in patients without a previous diagnosis of Barrett s esophagus. All these studies represent arguments in favour of a progression from milder to more severe forms of GERD occurring with time and against a hypothesis of segregation of subgroups, as proposed by some authors [8]. Moreover, a biological plausibility exists for this clinical progression, since the molecular abnormalities found in BE

4 412 F. Pace et al. / Digestive and Liver Disease 39 (2007) Fig. 1. Progression of oesophageal cells to cancer follows the metaplasia dysplasia adenocarcinoma sequence; characteristic alterations in cell survival and adhesion occur (from Ref. [31], with permission). patients are also progressive: as an example, an increased G1 fractions develop as an early event in the (nondysplastic) metaplastic epithelium due to chronic GE (gastroesophageal) reflux, but S phase and 4 N fractions are typically normal at this stage. As progression (inflammation due to uncontrolled reflux) continues, the p53 gene can be inactivated in diploid cells by mutation and 17p allelic loss, and a population of cells develops with increased 4 N (G2/tetraploid) fractions [31] (Fig. 1). Diploid cells may also develop 9p allelic loss, and with continued instability, clones of aneuploid cells eventually develop, with typical increase in S phase fractions, which lead to cancer [34] (Fig. 1). This is a progressive multistep process [35], similar to the cascade proposed by Correa et al. to explain the development of gastric cancer following various phases of increasingly severe chronic gastritis [36]. Finally, in the above quoted multinational European study called ProGERD, after 2 years, 45 patients had developed endoscopically and histologically confirmed Barrett s mucosa, equivalent to an annual incidence rate of 0.6%. However, the incidence rates were higher with increasing baseline severity of GERD: patients with Los Angeles grade C/D oesophagitis were 10 times more likely to develop confirmed Barrett s esophagus than NERD patients, and those with LA A/B were three times more likely to develop confirmed Barrett s esophagus [23]. 2. Discussion In 1993, Ollyo et al. [27] wrote the natural history of erosive reflux oesophagitis and its evolution are not clearly understood. Two models have been proposed in the literature, one diametrically opposed to the other. In the first model, oesophagitis evolves progressively from erosions to ulceration and stenosis. In the second model, the patients develop oesophagitis whose grade indicates the severity of the disease (erosion, ulcer, stenosis); the oesophagitis may heal but it will not progress to a more severe grade. While it is to be recognized that at the time no mention was made on the subgroup of NERD, which is a relatively newly defined segment of GERD spectrum, the controversy proposed by Ollyo is still open. In recent years, Fass et al. [8] have described the spectrum of GERD as composed of three unique patients groups not communicating with each other (NERD, erosive and Barrett s) and thus not progressive. To use Fass words, the evidence suggests that GERD patients belong to only one of these subgroups for the remainder of their lives. On the contrary, we think that there is significant movement in between these groups and have consequently discarded the above conceptual framework as essentially not supported by evidence or even wrong [7]. Rather, we propose that GERD patients be considered as a continuum, in which it is possible to fluctuate over time from one severity group to another, although we acknowledge that the data from natural history of the different patients subgroups are very limited. In this article, we have presented robust evidence that at least in a minority of GERD patients a progression occurs during the course of the disease, with, for example, mucosal erosions developing in a previously nonerosive GERD patient, or more severe damage occurring in patients with mild oesophagitis, or finally with a complication such as intestinal metaplasia or adenocarcinoma arising in a previously uncomplicated case. We acknowledge that this process possibly takes place only in some cases and is not the rule, but we interpret this finding as the consequence of early and aggressive treatment that most symptomatic GERD patients receive; on the contrary, a clear picture of how frequently a progression occurs in the real world (i.e. in the general population and not in tertiary referral centres) is lacking. Of course, the proof of the continuum existing in GERD spectrum does not mean that all the steps necessarily evolve in every patient, but rather that without appropriate therapy, uncomplicated cases may eventually become very severe. In reviewing the existing literature, it becomes clear that very few studies directly and prospectively address the issue

5 F. Pace et al. / Digestive and Liver Disease 39 (2007) of the natural history of GERD; in fact, most of the studies are retrospective in nature and are designed for other purposes. Moreover, the methodological problems affecting these studies are considerable, as already focused by others [25]. In our opinion, the most relevant problem is that very rarely if ever, have follow-up studies been conducted in the absence of active, antisecretory, therapy. This of course will interfere with the natural trend of GERD toward a progression, outshadowing this feature. One example comes from the recent, restrospective analysis of more than 2000 GERD patients conducted at the Hines Veterans Affairs Hospital endoscopy clinic in the United States [37]. In this study, the oesophageal mucosa was evaluated 7.6 years apart in patients taking continuous or intermittent antisecretory treatment; even under these circumstances, interestingly, in 11% of the sample the endoscopic appearance at second look was worse than at baseline. Independent of the possibility of flaws and biases in restrospective studies addressing the natural history of GERD, we do think that available studies allow us to model GERD as a continuum of different subgroups composing its spectrum, i.e. a non-categorial disease, as clearly shown by Labenz et al. [23]. 3. Conclusions We concluded our 2004 paper [7] as follows: In the end, what really matters is the concept that GERD is a multifaceted (spectrum) clinical problem, and that a modern approach should focus on the symptoms rather than on the presence of a 5-mm mucosal break in the distal esophagus. On the contrary, almost all available guidelines indicate that an endoscopy is only to be performed in particular circumstances (refractoriness to an empirical symptom-driven therapy, or presence of alarm features). From this perspective, the rigid division into three patients subgroups (NERD, erosive and Barrett s) appears to limit rather than expand our concentration on symptoms, because it implicitly calls for endoscopy in all patients at baseline (in order to categorize patients into one or another subgroup). In our opinion the very process of fragmenting the patient s spectrum (or iceberg) into non-communicating categories is arbitrary and not supported by clear evidence. On the contrary, annual progression rates from NERD to ERD is possibly as high as 30%, 22% is the maximal rate of annual progression of oesophagitis severity and 13% the rate of annual Barrett s esophagus development in ERD patients, according to a recent systematic review [25]. In conclusion, in a consistent minority of GERD patients, the disease is progressive with time. This may have significant consequences, as for example the need for periodic reassessment of disease severity, off treatment and/or the screening for Barrett s esophagus throughout the patient s life. More studies of better quality are needed for a further understanding of the natural history of GERD disease. Practice points The belief that GER disease might be divided into three categories, i.e. nonerosive GERD (or NERD), erosive GERD (or ERD) and Barrett s esophagus, not communicating with each other seems not to be proven by the current literature; and in fact, a progression from milder to more severe forms of GERD is both possible and documented. Reviewing the studies conducted on NERD patients indicate that about 10% of them per year develop erosive disease. As far as the patients with erosive oesophagitis are concerned, their annual progression rate is up to 22%. Finally, there is evidence that a proportion between 1% and 13% of patients with erosive oesophagitis developed Barrett s esophagus annually. It is however also possible that NERD or ERD patients might develop adenocarcinoma even in the absence of Barrett s esophagus. Research agenda To create a large, possibly national, database containing clinical, endoscopic and therapeutic informations of large cohort of GERD patients and to update it regularly. Conflict of interest statement None declared. List of abbreviations BE, Barrett s esophagus; GI, gastrointestinal; GE, gastroesophageal; EE, erosive esophagitis. References [1] Cantù P, Savojardo D, Carmagnola S, Penagini R. Impact on referral for gastro-oesophageal reflux disease on the workload of an academic Gastroenterology Unit. Dig Liver Dis 2005;37: [2] Stanghellini V, Cogliandro R, Cogliandro L, De Giorgio R, Barbara G, Corinaldesi R. Unsolved problems in the management of patients with gastroesophageal reflux disease. Dig Liver Dis 2003;35: [3] Howard PJ, Heading RC. Epidemiology of gastro-oesophageal reflux disease. World J Surg 1992;16:

6 414 F. Pace et al. / Digestive and Liver Disease 39 (2007) [4] El-Serag HB, Sonnenberg A. Associations between different forms of gastro-oesophageal reflux disease. Gut 1997;41: [5] Locke III GR, Talley NJ, Fett SL, Zinsmeister AR, Melton III LJ. Prevalence and clinical spectrum of gastro-oesophageal reflux: a population-based study in Olmsted County, Minnesota. Gastroenterology 1997;112: [6] Wong WM, Lai KC, Lam SK, Hui WM, Hu HC, Lam CLK, et al. Prevalence, clinical spectrum and health care utilization of gastrooesophageal reflux disease in a Chinese population: a population-based study. Aliment Pharmacol Ther 2003;18: [7] Pace F, Bianchi Porro G. Gastroesophageal reflux disease: a typical spectrum disease (a new conceptual framework is not needed). Am J Gastroenterol 2004;99: [8] Fass R, Ofman JJ. Gastroesophageal reflux disease: should we adopt a new conceptual framework? Am J Gastroenterol 2002;97: [9] Labenz J, Jaspersen D, Kulig M, Leodolter A, Lind T, Meyer-Sabelelk W, et al. Risk factor for erosive esopahgitis: a multivariate analysis based on the ProGERD study initiative. Am J Gastroenterol 2004;99: [10] Quigley EMM. Factors that influence therapeutic outcomes in symptomatic gastroesophageal reflux disease. Am J Gastroenterol 2003;98(Suppl.):S [11] Tack J, Fass R. Approaches to endoscopy-negative reflux disease: part of the GERD spectrum or a unique acid-related disorder? Aliment Pharmacol Ther 2004;19(Suppl. 1): [12] Malfertheiner P, Lind T, Willich S, Vieth M, Jaspersen D, Labenz J, et al. Prognostic influence of Barrett s oesophagus and Helicobacter pylori on healing of erosive gastro-oesophageal reflux disease (GORD) and symptom resolution in non-erosive GORD: report from the ProGORD study. Gut 2005;54: [13] Fass R. Distinct phenotypic presentations of gastroesophageal reflux disease: a new view on natural history. Dig Dis 2004;22: [14] DeVault KR, Castell DO. Updated guidelines for the diagnosis and treatment of gastroesophageal reflux disease. Am J Gastroenterol 2005;100: [15] Pace F, Manes G, Conio M, Bianchi Porro G. Pretreatment endoscopy Pro & Contra: endoscopy is needed before treatment in all patients with gastroesophageal reflux disease. Endoscopy 2006;38: [16] Pace F, Santalucia F, Bianchi Porro G. Natural history of gastrooesophageal reflux disease without oesophagitis. Gut 1991;32: [17] Pace F, Bollani S, Molteni P, Bianchi Porro G. Natural history of gastro-oesophageal reflux disease without oesophagitis (NERD) a reappraisal 10 years on. Dig Liver Dis 2004;36: [18] Schindlbeck NE, Klauser AG, Berghammer G, Londong W, Mueller- Lissner SA. Three years follow up of patients with gastroesophageal reflux. Gut 1992;33: [19] Isolauri J, Luostarinen M, Isolauri E, Reinikainen P, Viljakka M, Keyrilainen O. Natural course of gastroesophageal reflux disease: years follow-up of 60 patients. Am J Gastroenterol 1997;92: [20] McDougall NI, Johnston BT, Kee F, Collins JSA, McFarland RJ, Love AHG. Three to 4.5 year prospective study of prognostic indicators in gastrooesophageal reflux disease. Scand J Gastroenterol 1998;33: [21] Pace F, Pallotta S, Molteni P, Zentilin P, Russo L, Savarino V, et al. Natural history of NERD in 3 Italian tertiary referral centres after 5 years of follow-up. Gut 2006;55(Suppl. V):A62. [22] Bayerdorffer E, Sipponen P, Bigard M, Weiss W, Mearin F, Rodrigo L, et al. Esomeprazole 20 mg continuous vs. on demand treatment of patients with endoscopy-negative reflux disease (ENRD). Gut 2004;53(Suppl. 6):A106. [23] Labenz J, Nocon M, Lind T, Leodolter A, Jaspersen D, Meyer- Sabellek W, et al. Prospective follow-up data from the ProGERD study suggest that GERD is not a categorial disease. Am J Gastroenterol 2006;101: [24] Kawanishi M. Will symptomatic gastroesophageal reflux disease develop into reflux esophagitis? J Gastroenterol 2006;41: [25] Fullard M, Kang JY, Neild P, Poullis A, Maxwell JD. Systematic review: does gastro-oesophageal reflux disease progress? Aliment Pharmacol Ther 2006;24: [26] Manabe N, Yoshihara M, Sasaki A, Tanaka S, Haruma K, Chayama K. Clinical characteristics and natural history of patients with low-grade esophagitis. J Gastroenterol Hepatol 2002;17: [27] Ollyo JB, Monnier P, Fontolliet C, Savary M. The natural history, prevalence and incidence of reflux oesophagitis. Gullet 1993;3(Suppl.):3 10. [28] Cameron AJ, Lomboy CT. Barrett s esophagus: age, prevalence and extent of columnar epithelium. Gastroenterology 1992;103: [29] Bajbouj M, Reichenberger J, Neu B. A prospective multicenter clinical and endoscopic follow-up study of patients with gastroesophageal reflux disease. Z Gastroenterol 2005;43: [30] Lagergren J, Bergstrom R, Lindgren A, Nyren O. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med 1999;340: [31] Chow WH, Finkle WD, McLaughlin JK, Frankl H, Ziel HK, Fraumeni Jr JF. The relation of gastroesophageal reflux disease and its treatment to adenocarcinomas of the oesophagus and gastric cardia. JAMA 1995;274: [32] Farrow DC, Vaughan TL, Sweeney C, Gammon MD, Chow WH, Risch HA, et al. Gastroesophageal reflux disease, use of H2 receptor antagonists and risk of esophageal and gastric cancer. Cancer Causes Control 2000;11: [33] Conio M, Cameron AJ, Romero Y, Branch CD, Schleck CD, Burgart LJ, et al. Secular trends in the epidemiology and outcome of Barrett s oesophagus in Olmsted County, Minnesota. Gut 2001;48: [34] Reid BJ, Barrett MT, Galipeau PC, Sanchez CA, Neshat K, Cowan DS, et al. Barrett s esopahgus: ordering the events that lead to cancer. Eur J Cancer Prev 1996;5(Suppl. 2): [35] Jankowski JA, Harrison RF, Perry I, Balkwill F, Tselepis C. Barrett s metaplasia. Lancet 2000;356: [36] Correa P, Haenszel W, Cuello C, Tannenbaum S, Archer M. A model for gastric cancer epidemiology. Lancet 1975;2: [37] Sontag S, Sonneneberg A, Schnell TG, Leya J, Metz A. The long-term natural history of gastroesopahgeal reflux disease. J Clin Gastroenterol 2006;40: