Gastric Mucosa Cancer Risk: The Operative Link on Gastritis Assessment Staging System
Article information
Abstract
Gastric mucosa atrophy is the cancerization field for gastric cancer development. Atrophic changes in the gastric mucosa can be assessed endoscopically and histologically. Gastric serology (e.g., pepsinogens and their ratio, and gastrin) may provide information on the functional status of the mucosa. However, outside large-scale epidemiological studies, the clinical reliability of serology at the individual patient level remains controversial. Histology-based estimates of atrophy-associated neoplastic risk require mucosal specimens from both the antral/mucus-secreting and corpus/fundus oxyntic compartments. Inspired by the clinical–biological framework of cancer staging, the Operative Link on Gastritis Assessment (OLGA) staging system classifies the severity and topography of mucosal atrophy into five stages, ranging from 0 to IV. International trials have validated the stepwise increasing risk of cancer with disease progression. Based on this evidence, OLGA stages are classified into low-risk (0–I) and high-risk (III–IV), with stage II remaining clinically controversial. The prevalence of high-risk stages varies according to patient- and population-specific risk factors. Among these, the most significant is the prevalence of Helicobacter pylori, a recognized oncogenic agent. Bacterial infection at the time of the initial endoscopy requires eradication therapy regardless of the stage of gastritis. Histological stage reliably identifies the endoscopic population at a high risk of cancer progression and shapes the schedule of endoscopy/histology-based surveillance. Serology at the index endoscopy may circumvent the inherent inconsistencies of population-based studies. This patient-specific serology could serve as an interval for non-invasive monitoring within the endoscopy/histology follow-up framework.
INTRODUCTION
Under normal conditions, the lamina propria of the gastric mucosa hosts a few resident lymphocytes, histiocytes, rare plasmacytes, and extremely rare or no polymorphs [1]. Gastritis is defined as the microscopic presence of a leukocyte population exceeding the resident component.
Dyspepsia is the most common presentation of gastritis, ranging from minimal epigastric-centered discomfort to alarming symptoms such as vomiting or bleeding [2]. In such cases, esophagogastroduodenoscopy (EGD) is commonly recommended along with gastric mucosal biopsy.
Gastric serology includes testing of gastric mucosal functions (e.g., pepsinogens and gastrin), Helicobacter pylori status, and gastric autoimmunity. Pepsinogen 1 (Pg-1) is mainly secreted by the oxyntic glands, whereas Pg-2 is secreted by all the gastric and proximal duodenal Brunner’s glands. Pepsinogen secretion may be affected by inflammatory conditions (e.g., H. pylori infection) [3], atrophic-metaplastic transformation of the gastric lining, or both [4-9]. Gastrin 17 is the most abundant circulating gastrin peptide produced by neuroendocrine antral G cells targeting oxyntic enterochromaffin-like cells (ECL) and parietal cells and promotes acid secretion. Increased G17 levels are associated with gastric hypo- or achlorhydria [5]. Circulating antibodies may provide information on previous or current H. pylori-positive status. Clinical suspicion of autoimmune gastritis (AIG) can be confirmed by the serological detection of autoantibodies against anti-parietal cells and anti-intrinsic factors [10-14]. Active H. pylori infection can be confidently diagnosed using the urea breath test (13C-UBT) and H. pylori stool antigen detection [3,15-17].
The etiology is the most reliable reference for the classification of gastritis [18]. The World Health Organization distinguishes between communicable and noncommunicable etiological agents [19]. Both etiologies may result in self-limiting and non-self-limiting inflammatory lesions. Worldwide, longstanding (i.e., chronic), non-self-limiting gastritis is primarily caused by H. pylori infection [20]. The second most common cause is a cytolytic T cell-mediated attack on gastric parietal cells (AIG), with a significantly lower prevalence than that of H. pylori infection [21-24].
Long-standing gastric mucosal inflammation may lead to epigenetic and genetic molecular disarrays, impaired mucosal function, and microscopic disarrangement of the gastric lining. The combination of these inflammation-driven conditions creates a “field of cancerization” that can cause and enable the progression to gastric cancer (GC). The microscopic hallmark of this GC-promoting condition is the loss of native gastric glands, which is commonly recognized as “mucosal atrophy.” [25-27]
Irrespective of its etiology, the cancerization field topographically involves the gastric mucosa far beyond the focal area, where precancer or cancer lesions may become endoscopically/histologically detectable [28-31]. Any histological assessment of atrophy should comprehensively consider 1) the topographical extent of the cancer-prone mucosa and 2) the microscopic spectrum of atrophy phenotypes involved in cancer promotion.
This review focuses on the clinical priority of histological reporting of atrophy in gastric biopsy specimens. Histological staging of gastric mucosal atrophy provides clinicians with an easily accessible diagnostic information on the risk of GC tailored to patient-specific gastric conditions.
Twenty years ago, inspired by staging systems applied in oncology or inflammatory liver diseases [31], an international group of gastroenterologists and pathologists (Operative Link on Gastritis Assessment [OLGA]) proposed a framework (OLGA system) for staging gastric mucosal atrophy [32-34]. The disease stage was used to determine the severity of atrophy and its associated GC risk. Building on the original OLGA staging proposal, other staging systems have been suggested for the endoscopic or histological assessment of atrophy.
CLASSIFICATION OF GASTRITIS BASED ON ETIOLOGY, CLINICAL COURSE, MUCOSAL ATROPHY
More than a nosological exercise, the classification of gastritis provides clinically important information regarding diagnostic workup, treatment, and patient monitoring.
The etiology-based classification of gastritis includes environmental (i.e., communicable/transmissible) and host-related (non-communicable/non-transmissible) factors (Table 1). Based on the clinical presentation (i.e., symptoms and signs), gastritis is usually classified as either non-self-limited (commonly called “chronic” or “long-standing”) or self-limited. The latter is mostly characterized by sudden-onset and quickly remitting dyspeptic symptoms (clinically defined as “acute”). From a histological perspective, the cellular component of the inflammatory population may not correspond to the traditional, so-called “acute” (predominantly neutrophilic polymorphonuclear cells) versus “chronic” (mononuclear cells predominate) gastritis, and such a histological definition should be discouraged.
Histologically, gastritis is classified into two main categories: non-atrophic and atrophic, the latter being associated with an increased risk of GC [27,35]. Histological reporting of gastric mucosal atrophy requires distinguishing atrophic changes according to the gastric compartment in which they occur, specifically the corpus/fundus (oxyntic) or the antrum/angulus (mucus-secreting). Coexisting inflammatory lesions present a distinct diagnostic message, supported by the semantic distinction between “gastritis” and non-inflammatory gastric mucosal lesions. Lymphocytic/plasmacytic infiltrates should be graded semi-quantitatively as low or high-grade. The polymorphonuclear cell infiltrate should be reported, indicating whether detected in the lamina propria, glandular lumen, or inter-epithelial spaces (all these lesions are usually defined as “activity,” see below).
GASTRITIS: THE HISTOLOGICAL PHENOTYPES
Damage to the gastric mucosa can be classified into non-atrophic and atrophic types, based on the population of native glandular units. This dichotomous distinction influences the risk of GC. Both atrophic and non-atrophic mucosa may coexist with inflammatory lesions.
Atrophic transformation encompasses a spectrum of microscopic phenotypes that differ depending on the pathology of the gastric compartment in which they occur. A detailed description of the biology of atrophy is beyond the scope of this review; however, a few details on this “technical issue” may promote communication among pathologists, gastroenterologists, and medical practitioners.
Gastric mucosal atrophy is defined as the loss of gastric glands appropriate for the oxyntic (corpus/fundus) and mucus-secreting antral/pyloric stomach down to the incisura angularis.
The phenotypes of atrophy include: 1) shrinkage or complete disappearance of glandular units replaced by fibrotic lamina propria. This reduces the glandular mass without modifying the original cell phenotype; 2) replacement of native glands with metaplastic glands featuring a new commitment (i.e., intestinal [IM] and/or pseudo-pyloric metaplasia [PPM]). The number of glandular units is not necessarily lower, but the metaplastic replacement of the native glandular population results in fewer glandular structures being “appropriate” for the compartment concerned (i.e., oxyntic glandular units in the proximal stomach and mucus-secreting glands in the distal compartment, down to the incisura angularis). The metaplastic replacement of native glandular units results in the morphofunctional “loss of the appropriate glands.” There are two main types of gastric metaplasia.
PPM replaces a native population of oxyntic glands, which is consistent with the comprehensive definition of gastric mucosal atrophy [36-38]. The pseudopyloric epithelium is marked by the immunohistochemical expression of trefoil factor 2 [39]. The spotty expression of Pg-1 revealed native oxyntic commitment of the pseudo-pyloric glands [39]. Clinical and molecular studies have explored the potential of PPM in GC progression. Their conclusions can be summarized as follows: 1) in experimental models, the spasmolytic polypeptide-expressing metaplasia (SPEM)-related molecular profile has potential for cancer evolution [40]; 2) in humans, PPM can be confidently considered as the morphological counterpart of the SPEM’s molecular profile [41]; 3) the loss of oxyntic Pg-1-producing cells leads to the “PPM-atrophic subtype” [42]; and 4) experimental and clinical studies support the hypothesis that PPM may either undergo to further modification acquiring the phenotype of intestinal epithelia [43], or coexists with glands’ intestinalization [44,45].
Intestinal metaplasia (IM) may develop in native mucus-secreting (antral) epithelia or antral oxyntic glands (i.e., PPM). Different IM subtypes have been distinguished based on whether the metaplastic epithelium phenotype more closely resembles that of the colon (“incomplete” or colonic-type intestinal metaplasia; so-called Type I-III IM) or that of the small intestine (“complete,” Type I IM). Histologically proven colonic metaplasia identifies stomachs that are at risk of cancer development. Given the significant correlation between the score for the incomplete-type IM and its total extent, IM subtyping is primarily applied in research settings rather than in clinical practice [46]. Recent data have linked small intestinal-type IM to the atrophic pattern of AIG [47-49].
Globally, at least 90% of GC cases originate from atrophic gastritis. An increased risk of GC is also associated with successfully eradicated patients who harbor extensive (high-stage) cancer-promoting mucosal atrophy, which persists after H. pylori eradication [12,49-51]. In these patients, morpho-functional atrophic transformation results in a combination of cancer-promoting changes, including DNA toxicity, modifications of the cytokine cascade, oxidative damage, and cancer-prone remodulation of the native microbiota [52,53].
The chronological sequence and the oncogenic impact of these events in the GC “cascade” are only partially known. However, the occurrence of GC in patients who develop atrophic changes before bacterial eradication suggests an increased atrophy-specific cancer risk even after the infection is cured [54]. From a clinical perspective, this evidence supports the continued surveillance of patients with extensive atrophy, even after the successful eradication of H. pylori [51].
GASTRITIS OLGA STAGING IN CLINICAL PRACTICE
The clinical importance of OLGA staging lies in its ability to accurately predict GC risk. However, regardless of the stage of gastritis, all H. pylori-positive patients benefit from the bacterial eradication [2,12].
Table 2 outlines the histological criteria used to assess the OLGA stages. Based on their associated risk of cancer progression, stages 0, I, and II are considered to have a low GC risk, whereas Stages III and IV are associated with an increased risk of cancer progression. Histological assessment of the OLGA stage involves the following steps: 1) obtain a biopsy set compliant with the Houston-Updated Sydney system, representative of the oxyntic and mucus-secreting gastric compartments; 2) submit the biopsy samples in at least two separate vials, each labeled according to the compartment (antrum/incisura angularis versus corpus fundus) from which they were obtained. Any focal lesion must be targeted by an additional biopsy sampling t and submitted to separate topographically labeled vials; 3) comprehensively score of all the atrophy subtypes by the gastric mucosal compartment; and 4) combine atrophy scores from the corpus/fundus and antral/angular compartments and combine them in the final stage, ranging from 0 to IV.
In some cases, due to the inconsistency in sampling or tissue orientation, OLGA staging may be deferred.
The first two actions are the responsibility of the endoscopist, while steps three and four fall under the responsibility of the pathologist.
Steps 1 & 2: based on embryonic patterning, the stomach is described as “two organs in one sac.” [55] This dual mucosal commitment leads to distinct pathophysiological changes in the oxyntic/proximal and the mucus-secreting/distal stomach mucosa. The two main models of atrophy highlight the diversity of the gastric mucosa.
In the natural history of H. pylori infection, inflammation and atrophy initially affect the mucus-secreting mucosa, progressing to the oxyntic stomach as the disease advances [56,57].
In AIG, lesions are restricted to the oxyntic stomach [43,58]. The diagnostic impact of the topographic distribution of atrophy results in three major recommendations (Table 2): a) obtain two biopsy samples from the oxyntic mucosa, one targeting the anterior wall and one targeting the greatest curvature; b) obtain a biopsy sample from the transitional mucosa between the corpus and antral mucosa (incisura angularis zone along the lesser curvature); and c) obtain two biopsy samples from the mucus-secreting/distal stomach down the incisura angularis. The samples identified as “a” and those “b and c” should be submitted in separate vials, labeled by biopsy site. Distinguishing “a” from “b” biopsy specimens would be desirable. The sampling protocol must include any occasional focal lesion(s), which should be submitted in a separate vial and identified by its/their mucosal topography.
Steps 3 & 4 cover the microscopic assessment of atrophic/inflammatory lesions and the formatting of the diagnostic report. While the histological assessment of the atrophy score is performed by pathologists, gastroenterologists and endoscopists should be aware of the extent to which the quality of biopsy sampling affects the reliability of staging. The atrophy grade is assessed microscopically in biopsy samples from the overall antral/angular biopsy set and the corpus/fundus specimens. The atrophy score encompasses all phenotypic changes (non-metaplastic and metaplastic [IM and PPM]). In both the oxyntic and mucus-secreting compartments, the score values range from zero (no atrophy: grade=0) to more than 60% (atrophy score grade=3; Table 2). The combination of atrophy grades from the two gastric compartments determined the OLGA stage, which ranged from 0 to IV (Table 2). The atrophy stage is independent of the assessment of any coexisting inflammatory component (i.e., gastritis), which represents an adjunctive diagnostic parameter. In each compartment, inflammation was scored as absent, low, or high. The explicit distinction between inflammation and atrophy scores is crucial in H. pylori-associated diseases, where minimal inflammation may persist after the successful eradication of H. pylori [12,22,27].
In routine diagnostic practice, high-grade inflammatory lesions (primarily associated with Helicobacter pylori infection) or inconsistent orientation of biopsy specimens can present challenges in the assessment of atrophy. This “indefinite for atrophy” phenotype should prompt the deferral of the OLGA stage assessment until post-eradication follow-up or until a new set of biopsies is obtained, when clinically indicated [27].
Based on this, the diagnostic report should include [21]: 1) the distinction between non-atrophic and atrophic mucosal damage; 2) reporting the atrophy grades (e.g., G0, G1, G2, and G3), distinguishing the corpus/fundus and antrum/angulus biopsies (e.g., antral atrophy score=Gx; corpus/fundus atrophy score=Gx), and their resulting OLGA stage (Fig. 1, Table 2); 3) the distinct score of inflammation and gastritis “activity”; and 4) a plausible etiological hypothesis based on the gastritis histological phenotype.
Obtaining gastric biopsy samples for diagnostic purposes. The stomach is opened along the greater curvature. Biopsy sampling protocol suggested by the updated Sydney System (left). The system recommends obtaining biopsy samples from the antral (S1–S2), angular (S3), and oxyntic (S4–S5) mucosa. According to OLGA staging, the histological severity of overall atrophic changes is scored as the mean atrophy percentage in the antral/incisural (S1–S3) and body/fundus mucosa (S4–S5). In Eastern Asian clinical practice, the assessment of gastric atrophy primarily focuses on the cranial movement of the “atrophic border.” This border marks the transition between the most caudal oxyntic mucosa and the most cranial mucosecreting mucosa, as described by Kimura–Takemoto.29 The endoscopic severity of atrophy is ranked as mild (C1, C2), moderate (C3, O1), and severe (O2, O3); the most informative biopsy samples should be obtained from sites A, B, C, and D (on the right). OLGA, Operative Link on Gastritis Assessment.
Capelle et al. [32] suggested an alternative staging format. Their proposal replaces the original OLGA-atrophy score (including non-metaplastic, and metaplastic subtypes) with a simplified scoring system limited to IM. By reducing the histological variables considered by the OLGA-staging, the Operative Link on Gastric Intestinal Metaplasia assessment (OLGIM) system has been reported to achieve higher interobserver consistency, while potentially affecting the accuracy of GC risk estimates [32,59-62].
The Operative Link on Gastric Intestinal metaplasia and Glandular Atrophy Assessment (OLGIMA) system is the most recently proposed staging system [63]. By excluding PPM from the atrophy score, staging relies exclusively on non-metaplastic and IM atrophic changes. As for the OLGA and OLGIM systems, OLGIMA stages III and IV were associated with increased GC risk. The reproducibility and clinical significance of the OLGIMA system have not been evaluated outside the proposing centers.
The similarities and differences between the OLGA and OLGIM staging systems can be summarized as follows.
- Both systems use a Sydney-based protocol for biopsy sampling and recommend submitting the specimens separate vials [64].
- Both systems recognize histologically assessed mucosal atrophy as a key factor affecting GC risk estimates; however, they differ in their definition of atrophy.
- Both systems require structured training and operator expertise [65].
- Both systems aim to convey a straightforward message to guide patient management. The differences in the phenotypic variables included in the atrophy scores of OLGA and OLGIM lead to distinct advantages: 1) OLGIM’s IM-restricted atrophy assessment increases inter-observer consistency; 2) OLGA’s inclusion of a broader spectrum of atrophy phenotypes expands the estimates of high-risk gastritis stages [66]. In a retrospective analysis of 4556 cases, 5.8% of high-risk OLGA stages were reclassified as low risk according to the OLGIM criteria. Isajevs and colleagues compared the interobserver agreement and prognostic performance of the OLGA and OLGIM systems in a cohort of 836 patients. The authors’ conclusions were as follows: “OLGIM staging provided the highest interobserver agreement, but a substantial proportion of potentially high-risk individuals would be missed if only OLGIM staging is applied.” [60,67] Other studies comparing case distributions between the two systems found that a substantial number of cases were staged lower by OLGA than by OLGIM, which is surprising given the broader spectrum of atrophic changes included in the OLGA atrophy score [68-70].
- Both systems significantly predict the risk of metachronous GC after endoscopic submucosal dissection for early GC [71].
- Neither system includes information about coexisting inflammation, which should be graded separately.
OLGA STAGING IN CANCER RISK PREDICTION
International prospective and retrospective studies conducted in various epidemiological settings have supported the different cancer risks associated with each OLGA stage. Table 3 summarizes the results of two long-term follow-up studies involving 9191 consecutive endoscopy patients [72,73]. In follow-ups >50 months, both studies found that OLGA stages III and IV (i.e., high-risk stages) were associated with a significantly increased risk of developing neoplastic lesions. Conversely, the same studies reported a significantly lower risk of neoplasms, including intraepithelial neoplasia or invasive GC, in stages 0-I-II (low-risk stages).
Risk of incident gastric neoplasia (either non-invasive [synonym intraepithelial neoplasia] or invasive adenocarcinoma [GC]) by gastritis OLGA staging
In a study involving 7436 consecutive patients [72], multivariate analysis, including sex, age, OLGA stage, and H. pylori status at presentation, showed only the histological stage significantly linked to neoplastic progression (hazard ratios: Stage O=1; Stage I=12.7 [1.14–140.7]; Stage II=54.9 [5.63–534.6]; Stage III=712.4 [92.5–5484.5]; Stage IV: 1450 [166.7–12626]). Based on the above-mentioned experience, the low prevalence of stages III and IV (2.3% in endoscopic cohorts, increasing to 6.6% when stage II is included) supports the feasibility of effectively implementing secondary prevention strategies that combine esophagogastroduodenoscopy with appropriate biopsy sampling [16,74]. For H. pylori-negative patients (either naïve or previously eradicated), Fig. 2 provides a timing framework for endoscopic and histological surveillance based on the OLGA stage at enrollment. Stricter follow-up may be warranted for at-risk endoscopic lesions, as well as for clinical variables such as age, sex, family history of gastric cancer, and biological risk factors such as a predominance of incomplete-type IM [16].
Prevalence of OLGA stages (0-IV) in an endoscopic population recruited in an epidemiological study with low to medium gastric cancer risk. The distribution of patients by stage (percentage) is represented by the length of the bars. Surveillance (including endoscopy and histological follow-up) is guided by histology in H. pylori-negative patients (whether naïve or eradicated) according to OLGA stage. This occurs when there are no accompanying endoscopic, demographic (such as sex, age, or family history of gastric cancer), or biological risk factors (such as the predominance of incomplete-type intestinal metaplasia, etc.). OLGA, Operative Link on Gastritis Assessment.
These findings from long-term EGD/histology follow-up suggest:
- Irrespective of the histological stage at enrollment, H. pylori-positive status requires treating for infection [2,75].
- In low-risk OLGA stages, evidence of widespread mucosal intestinalization on EGD [16], as well as the open-type Takemoto’s atrophy patterns [76], or any focal lesions (as detected by “conventional” white-light or by any other more sophisticated optical imaging technologies) requires extending the diagnostic workup [77,78].
- In patients with OLGA stage II (antral atrophy: G=0; oxyntic atrophy: G2 [i.e., oxyntic-restricted atrophy and inflammation]), the etiopathogenetic hypothesis of AIG can be plausibly suggested [43]. This hypothesis can be further substantiated by the concomitant observation of ECL hyperplasia. Such a gastritis phenotype should prompt the following actions: 1) including the histology-based etiological hypothesis in the histology report; 2) recommending serological testing for gastric (and extra-gastric) autoimmunity, hematological disorders (such as pernicious anemia); 3) suggesting endoscopic duodenoscopy (EGD) or histology surveillance that encompasses both the disease-specific risk of Type I neuroendocrine tumors and the infrequent occurrence of epithelial precancer or cancer lesions; and 4) investigating AIG-associated comorbidities, particularly inflammatory and neoplastic thyroid conditions [43].
THE SEROLOGICAL COUNTERPART OF GASTRIC MUCOSA ATROPHY
For at least three decades, “functional” gastric serology (i.e., Pg-1, Pg-2, Pg-1:Pg-2, and gastrin) has been proposed as a non-invasive and inexpensive method to identify subjects at increased risk for GC. Large-scale epidemiological studies have shown a strong correlation between mucosal atrophy and the pepsinogen ratio. Based on this evidence, Pg-serology has been proposed as a screening tool for patients who require further invasive diagnostic procedures [79-81]. However, beyond epidemiology, the clinical reliability of serology for ranking the severity of atrophy in individual patients remains controversial.
The limited use of serology in diagnostic practice is affected by significant biases, including demographic and lifestyle variables, local H. pylori epidemiology, current or prior H. pylori status, and ongoing therapies [82-85]. To minimize the confounding effect of these variables, the patient’s specific serology could be tested along with histology at index endoscopy. This serology could serve as a reference for a multimodal approach to atrophy monitoring that may replace non-invasive and invasive (more expensive EGD with biopsy) surveillance [86,87].
CONCLUSIONS
In Western countries, fewer than 30% of patients with GC survive for more than 3 years after diagnosis. This unacceptable mortality rate is due to neoplastic dissemination that has already occurred by the time GC is detected [88].
A global analysis by the International Agency for Research on Cancer (IARC) has recently reported that up to four in ten cancer cases worldwide (approximately 7.1 million cases) could be prevented. Gastric malignancies rank second among the most prevalent cancers worldwide [89], accounting for nearly half of all preventable cancers in both sexes. H. pylori eradication programs and cancer screening strategies have demonstrated efficacy in East Asian epidemiological settings, particularly in Korea and Japan. These initiatives have achieved significant success and anticipate even more promising outcomes [90-95].
To gain insights from the Asian model, the GC secondary prevention task force should include healthcare decision makers, general practitioners, gastroenterologists, endoscopists, and pathologists [17,96,97]. In this collaborative effort, pathologists are required to generate diagnostic messages that assess personalized GC risk, thereby providing reliable information for tailoring surveillance protocols [2,16,98].
Notes
Availability of Data and Material
Data sharing not applicable to this article as no datasets were generated or analyzed during the study.
Conflicts of Interest
Massimo Rugge, a contributing editor of the Korean Journal of Helicobacter and Upper Gastrointestinal Research, was not involved in the editorial evaluation or decision to publish this article. All remaining authors have declared no conflicts of interest.
Funding Statement
None
Acknowledgements
None
Authors’ Contribution
Conceptualization: Massimo Rugge. Data curation: Giulia Fiorini, Marta Sbaraglia. Methodology: Massimo Rugge, Robert M Genta, Gavino Faa. Writing—original draft: Massimo Rugge. Writin—review & editing: Massimo Rugge, Robert M Genta. Approval of final manuscript: all authors.
