The Sydney System for classification of gastritis 20 years ago


  • Conflict of interest
    Dr Pentti Sipponen is a scientific adviser of Biohit Plc, a Finnish company which develops and markets laboratory pipettes and laboratory tests. Over the years Dr Sipponen has been a lecturer, usually on gastritis, atrophic gastritis and gastric cancer, in several congresses, symposia and meetings arranged by many pharmaceutical companies (Gist-Brocades, AstraZeneca, Glaxo, Altana Pharma, Orion Pharma, Takeda, etc.). Dr Price has no conflicts of interest.

Dr Pentti Sipponen, Repolar Oy; Box 26, 02101 Espoo, Finland. Email:


The roots of research into gastritis go back into the early decades of the 20th century. Modern aspects of its classification and knowledge of its biological course and consequences were relatively well known even at the time that Helicobcter pylori was discovered by Robin Warren and Barry Marshall in 1982. This discovery, however, significantly changed the field, establishing that the commonest form of gastritis is simply an infectious disease, a finding that raised enormous interest in the subject amongst gastroenterologists, microbiologists, pathologists and basic researchers. However, many of these “new” players in the field often had a limited knowledge of the morphological aspects of gastric inflammations and chronic gastritis. As a consequence in the late 1980's a Working Party was set up to review the biology and natural course of chronic gastritis, to propose a new classification for gastritis, and to provide simple guidelines for reporting the pathology of gastritis in endoscopic biopsies in an attempt to bring uniformity to the subject and facilitate comparative studies in what was to be an era of high research activity. These guidelines, The Sydney System: A New Classification of Gastritis was presented to the World Congress of Gastroenterology in Sydney in 1990, and was later published as six papers in the Journal of Gastroenterology and Hepatology. Now, twenty years on, this review looks back on the birth of Sydney System and why it is still important and successful.

Twenty years ago, at the World Congress of Gastroenterology in Sydney in 1990, a working party presented the Sydney System: A New Classification of Gastritis which was subsequently published as six papers in the Journal Gastroenterology and Hepatology.1–6 These encompassed the pathology, the endoscopic aspects, the microbiology, autoimmunity and epidemiology of chronic gastritis.

The System was a major focus of attention at the Sydney congress and gained even more attention afterwards. The Working Party presentation had been carefully prepared in many pre-meetings, and the whole clinico-pathological consensus process was orchestrated by two initiators Professors George Misiewicz and Guido Tytgat. A Dutch pharmaceutical company, Gist-Brocades, kindly provided financial support that facilitated the numerous preparatory pre-Sydney World Congress meetings that were the essential basis for the final success. In those days the company marketed bismuth as a drug for the treatment of Helicobacter pylori—an option that still is valid.

Why was the Sydney System so important?

Even though a considerable amount was already known about gastritis itself, and about its natural course and disease associations, after the discovery of H. pylori by Robin Warren and Barry Marshall in 1982,7 the approach to the understanding of gastritis and upper gastrointestinal disease changed markedly. It became evident that H. pylori infection was the key factor and initiator of the majority of the pathology related to gastritis and its sequelae. Consequently interest in gastritis and the morphological appearances of endoscopic biopsies soared. Many physicians, microbiologists and basic researchers entered the field leading to a proliferation of international meetings and publications. Not unexpectedly with so many disciplines involved there was considerable diversity in approach and in terminology. This led to some basic misunderstandings at a clinical and pathological level. Therefore, the ultimate target of the Sydney System Working Party was to promulgate a common language applicable to the new knowledge about the biology and natural course of chronic gastritis, and also to provide simple guidelines for the documentation of the microscopical appearances in biopsy specimens together with an easily understandable classification for clinical and research purposes. An additional goal was also to provide guidelines for the reporting and classification of the endoscopic appearances of the gastric mucosa.3 The latter (“The Endoscopic Division” of the Sydney System) met with less success and this section of the System has remained in the shadow of the morphological guidelines (“The Histological Division”).

How was it done?

During 1988–1990, several pre-meetings of the Working Party were convened under the leadership of George Misiewicz and Guido Tytgat. Professor Stewart Goodwin represented microbiology in the group, and two pathologists, Ashley Price from London and Pentti Sipponen from Helsinki, spoke for histopathology. Behind the latter two were an important advisory and supportive group of European pathologists (Dr V Bogomoletz, Prof M Dixon, Prof J Haot, Prof K Heilmann, Prof E Solcia, and Prof M Stolte). The main Working Party representative for the clinicians and endoscopists was Professor Robert Strickland from Albuquerque. Noteworthy, Robert Strickland brought into the discussions of the Working Party his deep knowledge of the old research on various types of chronic atrophic gastritis. Already in 1973, when still living in Australia, Strickland described with McKay the atrophic gastritis of types A and B, the “A type” indicating corpus limited atrophic gastritis of autoimmune origin, and the “B type” representing the atrophic gastritis that is now known to be the consequence of gastritis related to H. pylori infection.8 This simple classification formed one of the bases of the Sydney System.

In practice, two “Sydney Working Party” groups existed working together and arranging the pre-meetings in the same place and at the same time. One group was the pathological the other mainly clinical. Within the latter's remit, aside from endoscopy, was to consider new options for the treatment of H. pylori. Guido Tytgat (chairman), David Graham, Adrian Lee, Barry Marshall, Michael Dixon and Anthony Axon were the members of this group.

What were the key issues in the Sydney System?

The Histological Division of the Sydney System was intended to be a practical guideline highlighting which of the morphological features of gastritis in endoscopic biopsies should be documented, and how these might be graded. The final pathology report would then convey the type, severity and extent of the gastric pathology linked to the etiology where possible. A single chart was designed on which to record the key parameters and be the quantitative basis for comparisons between biopsies from individual patients and between patient groups in therapeutic trials (Fig. 1).

Figure 1.

The chart designed for the histological division of the original Sydney System as presented to the Sydney World Congress of Gastroenterology in 1990, and published in the Journal of Gastroenterology and Hepatology in 1991. It incorporates etiology, topography and the morphological features to be documented when reading and reporting endoscopic gastric biopsies. The topography of gastritis is the core of the classification. Etiological hints can be added as a prefix and the graded variables as suffixes. Typical examples would be: “H. pylori pangastritis, severely active with mild panatrophy”, “Autoimmune corpus gastritis with severe atrophy and intestinal metaplasia”; “Reactive mild antral gastritis; inactive; no H. pylori”, etc.

The topography of the gastritis was considered the core of the classification. Succinctly this was gastritis restricted to the antrum, restricted to the corpus, or a pangastritis. The etiology of gastritis, if known, was to be added as a prefix (e.g. H. pylori antral gastritis; autoimmune corpus gastritis, etc.). As suffix, phrases any of five key graded morphological variables were to be included. These were1 chronic inflammation (chronic gastritis)2 the activity of the gastritis measured by the presence of polymorphonuclear leucocytes alongside the mononuclear inflammatory infiltrate3 intestinal metaplasia (IM)4 atrophy manifest by the loss of the normal mucosal glands, and5 the presence of H. pylori organisms. The guidelines recommended these five parameters were recorded separately for both antrum and corpus with at least two random biopsies to be taken from each site. Furthermore, it was recommended these parameters were to be semi-quantitatively graded as absent, mild, moderate or severe, each successive grade to represent an increase in severity of approximately one third.

The System provided a clear picture of the extent and topography of the gastritis and also its severity. In clinical diagnostic practice by adopting etiological prefix phrases, the core topography and morphological suffix phrases the histology report conveyed in a compact standard style the key data for that biopsy episode with a semi-quantitative format for future comparative episodes or studies. For example a report summary might read “H. pylori pangastritis, severely active with moderate antral atrophy and intestinal metaplasia, or “Autoimmune corpus gastritis with severe atrophy; no intestinal metaplasia”, etc.

The principles of classification for gastritis in the Sydney System, and the selection of the morphological key variables were based on the available scientific knowledge and on relevant papers published in the literature. Some of these basic backbone papers were the publications of Schindler in 1947 in which he described a “superficial gastritis” that may progress to atrophic gastritis with time.9 This description of the natural course and time-dependent worsening of chronic gastritis was further based on many reports and studies from Finland and Estonia. These indicating that up to one half of patients with H. pylori gastritis may get atrophic gastritis of some morphological type and grade during a lifetime.6,10–13 Correspondingly, it was realized that the complications and disease associations of H. pylori gastritis were dependent on the extent, topography and severity of the infection; e.g. the gastric cancer risk increased exponentially with the progression of gastritis towards an achlorhydric or hypochlorhydric stomach with severe mucosal atrophy and intestinal metaplasia.

The earlier studies of Strickland and McKay proposed the topographic classification of atrophic gastritis into types A and B which constituted the core issue of the System. The A type indicating a corpus limited atrophic gastritis of autoimmune origin in patients with achlorhydria and with serological signs of the autoimmunity.8 Type B indicated atrophic gastritis without an autoimmune background and with atrophy limited to the antrum. This B subtype was considered “idiopathic” but was, after the discovery of H. pylori, realized to be typically related to H. pylori infection. It corresponds with the type of atrophic gastritis considered multifocal (MAG; multifocal atrophic gastritis) by Pelayo Correa in his reports from the 80s,14 or atrophic gastritis of the AB type by Glass and Pitchumoni from the 70s.15 In 1988, gastritis of “C type” was proposed by Wyatt and Dixon indicating reactive gastric lesions that were neither H. pylori associated nor autoimmune, and believed to be “chemical” (reactive) or drug induced in nature.16

An important morphological parameter incorporated into the Sydney System from the literature was the “activity” of the gastritis. The activity of gastritis was first proposed and emphasized by Whitehead, Truelove and Gear in 1972.17 In their classification of gastritis, the Schindler—Glass-Pitchumoni—type topographical pattern of chronic gastritis was accepted but was complemented with the concept of the “activity”. The “activity” indicated the co-existence of polymorphonuclear inflammation alongside the mononuclear one. This phenomenon we now consider to reflect the reaction of the host against the infection and to be strongly associated with the risk of the progression of gastritis to an atrophic pattern which is also related to the acquisitions of cytotoxic H. pylori strains.

It is noteworthy that the guidelines presented in the Sydney System are still valid and applicable some 20 years later. In our view this is because the Sydney System was an appropriate synthesis at the time of the old knowledge on chronic gastritis mentioned above with the new data and knowledges about H. pylori acquisition and its consequences.

It must be said the Sydney System was not a classification in strict terms but more a practical guideline that provided what we hoped might become a flexible universal gastritis reporting grid. It allowed for a standard rapid documentation of the extent and severity of the gastritis at the time of examination of the patient. H. pylori gastritis is, however, a dynamic and progressive process in which the histopathological appearances will change with time (over years and decades), and the later stages are difficult to predict from a single set of biopsies. A precise forecast of the end point of chronic gastritis from a single set of biopsy specimens cannot be made. For example, one cannot say with any confidence whether a 30-year old patient with a non-atrophic H. pylori gastritis will or will not progress to develop an atrophic gastritis, and if it were the case which morphological and topographical type (plus or minus intestinal metaplasia) might ensue decades later.

What happened later on?

Following the World Congress presentation and publication of the Sydney System some prominent American pathologists expressed their disquiet. They considered it a “European” enterprise. Correa and Yardley, and Rubin, criticized the System in Gastroenterology18,19 for failing to take account of all the gastritides, and considered it was not a classification per se. However as already pointed out the goal of the Sydney System was not to be a textbook of gastric pathology but designed to encourage a standard methodology for the reporting of the appearances of the H. pylori gastritis and its consequences. Following the American initiative, a new two-day consensus meeting was arranged in Houston in 1994, after which another consensus report, the “Up-Dated Sydney System”, was published in 1997 by Dixon, Genta and Correa.20 In practice, in our opinion, this up-dated system merely added the recommendation to include biopsies from the angulus of the stomach (the presence of intestinal metaplasia in the angulus may be an early sign of atrophic multifocal gastritis), and it provided a helpful “visual analogue scale” for the grading of the histological parameters (chronic inflammation, activity, atrophy, intestinal metaplasia and H. pylori) that had been already listed in the original Sydney System.

The future aspects

Both the original Sydney System and the Up-Dated version are still acceptable guidelines and are in use in many centers throughout the world. Adoption of the guidelines in everyday practice has enhanced evidence-based medicine in understanding the dynamics and management of patients with chronic gastritis and highlighted the fact that the interpretation of gastric mucosal pathology by endoscopy alone is not reliable, and not good clinical practice.

Current and future research in the field of chronic gastritis correlated with the biopsy appearances should permit a more accurate forecast of outcome than is possible with the Sydney System. For example, techniques that elucidate alterations in signaling pathways or changes in the molecular pathology of the gastric epithelium to be visualized on tissue sections may provide such tools. In addition a non-invasive approach might be through the assay of the various gastric biomarkers from tissue fluids and blood plasma.21