AIM: To compare the reliability of gastritis staging systems in ranking gastritis-associated cancer risk in a large series of consecutive patients.

METHODS: Gastric mucosal atrophy is the precancerous condition in which intestinal-type gastric cancer (GC) most frequently develops. The operative link for gastritis assessment (OLGA) staging system ranks the GC risk according to both the topography and the severity of gastric atrophy (as assessed histologically on the basis of the Sydney protocol for gastric mucosal biopsy). Both cross-sectional and long-term follow-up trials have consistently associated OLGA stages III-IV with a higher risk of GC. A recently-proposed modification of the OLGA staging system (OLGIM) basically incorporates the OLGA frame, but replaces the atrophy score with an assessment of intestinal metaplasia (IM) alone. A series of 4552 consecutive biopsy sets (2007-2009) was retrieved and reassessed according to both the OLGA and the OLGIM staging systems. A set of at least 5 biopsy samples was available for all the cases considered.

RESULTS: In 4460 of 4552 cases (98.0%), both the high-risk stages (III + IV) and the low-risk stages (0 +I + II) were assessed applying the OLGA and OLGIM criteria. Among the 243 OLGA high-risk stages, 14 (5.8%) were down-staged to a low risk using OLGIM. The 67 (1.5%) incidentally-found neoplastic lesions (intraepithelial or invasive) were consistently associated with high-risk stages, as assessed by both OLGA and OLGIM (P < 0.001 for both). Two of 34 intestinal-type GCs coexisting with a high-risk OLGA stage (stage III) were associated with a low-risk OLGIM stage (stage II).

CONCLUSION: Gastritis staging systems (both OLGA and OLGIM) convey prognostically important information on the gastritis-associated cancer risk. Because of its clinical impact, the stage of gastritis should be included as a conclusive message in the gastritis histology report. Since it focuses on IM alone, OLGIM staging is less sensitive than OLGA staging in the identification of patients at high risk of gastric cancer.

Gastric mucosal atrophy is by far the greatest risk factor for non-hereditary, intestinal-type distal gastric cancer (GC)[1]. The gold standard for atrophy assessment is histology, but non-invasive tests (mainly pepsinogen serology) are also applied for this purpose[1-3].

According to the current international literature, atrophy is defined as the “loss of appropriate glands”. This definition covers both the “loss” of native glands (replaced by fibrosis) and the metaplastic replacement of the appropriate (native) glands due to antral intestinalization, corpus antralization [i.e., spasmolytic polypeptide-expressing metaplasia (SPEM)] and/or intestinalization[2].

Consistent evidence correlates the extent/topography of atrophy with the risk of GC, and it is on these grounds that a system for staging gastritis [the operative link for gastritis assessment (OLGA) staging system] was proposed[4]. Gastritis stages (0 to IV) express increasing extents of atrophy, as assessed histologically on antral and corpus biopsies. In different epidemiological settings, both cross-sectional and long-term follow-up studies have consistently allocated a small minority of gastritis patients to stages III-IV, associating only this population with a significantly higher GC risk (high-risk OLGA stages)[3,5-10]. OLGA stages III-IV have also been consistently associated with molecular tissue markers of high-risk gastritis[11,12]. These correlations potentially support the advisability of endoscopic follow-up for such high-stage patients.

A significant correlation has been demonstrated between high-risk OLGA stages and pepsinogen serology; this correlation between organic and functional disease supports the rationale for implementing serology in GC secondary prevention programs[5].

A recently-proposed modification of the OLGA staging system (OLGIM)[10] basically incorporates the same staging frame, but replaces the “global” score for atrophy (in its different phenotypic variants) with the histological assessment of intestinal metaplasia (IM) alone. The rationale behind the OLGIM proposal lies in the fact that IM is easier to assess histologically than the “global” spectrum of the atrophic lesions (as in the OLGA approach).

This study compares the OLGA and OLGIM staging systems in the assessment of gastritis-associated gastric cancer risk (i.e., stages 0-I-II = low-risk stages vs Stages III-IV = high-risk stages).

Overall, 4552 consecutive biopsy sets were considered. The male/female ratio was 1/1.18, and the patients’ mean age was 55.1 years (median 57.0, range 20-89). For the males, the mean age was 55.0 years (range 20-88), while for the females it was 55.1 years (range 20-89).

Overall, 2967 biopsy sets (65.2%) showed no atrophic changes (i.e., stage 0 gastritis according to both OLGA & OLGIM) and the prevalence of the low-risk stages was 94.7% and 95.0% according to OLGA and OLGIM, respectively (Table 1).

In all, there were 67/4552 (1.5%) incidentally-found neoplastic lesions (either intraepithelial or invasive), including: 21 cases of LG-IEN; 6 cases of HG-IEN; 40 cases of GC (6 cases of diffuse-type proximal GC; 34 of intestinal-type distal GC). The M/F ratio among the neoplastic patients was 1.03/1 and their mean age (67.8 years; median 68; range 45-86) was significantly higher than that of the non-neoplastic patients (mean 54.8 years; median 57 years; range 20-89) (W; P < 0.001).

H. pylori was assessed histologically in 1698 patients (37.3%). Its prevalence among the neoplastic cases was 26 (38.8%), disregarding any previous eradication therapies. However, this study only focuses on the neoplastic risk associated with the stage of gastritis, without any specific reference to, or speculation about, the impact of the etiology on the morphogenesis of the gastric disease.

For 4460 out of 4552 cases (97.98%), low-risk stages (0 +I+ II) and high-risk stages (III + IV) were staged consistently using either OLGA or OLGIM criteria (Figure 1). For the 92 (2.0%) cases staged inconsistently, 14 were considered as low-risk using the OLGIM criteria and as high-risk according to OLGA. No cases staged as high-risk by OLGIM were down-staged when the OLGA criteria were applied.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Distribution of patients by the two considered staging systems (operative link for gastritis assessment vs operative link on intestinal metaplasia). OLGA: Operative link for gastritis assessment; OLGIM: Operative link on intestinal metaplasia.

The number of patients with high-risk stages (III-IV) was 243 according to OLGA and 229 according to OLGIM, i.e., among the 243 OLGA high-risk stages, 16 (6.6%) were down-staged by OLGIM; in particular, 14 OLGA stages III and IV were classified as stage II according to OLGIM.

In all, 67 intraepithelial (i.e., non-invasive) or invasive neoplastic lesions were detected. All the 27 intraepithelial neoplasia coexisted with intestinalized glands. Among the 40 cases of invasive adenocarcinoma, 6 (15%) were located in the cranial stomach and histologically featured a solid/diffuse-type GC; the other 34 (85%) were cases of intestinal-type GC. After distinguishing between low- and high-risk stages, a significant association emerged between stages III-IV and both intraepithelial and invasive neoplasia according to both the staging systems [W; P < 0.001 for both (Table 1)].

Fifty-nine of 67 (88.1%) and 57/67 (85.1%) intraepithelial or invasive neoplastic lesions were associated with high-risk OLGA and OLGIM stages, respectively. Six gastric cancers were detected in cases classified as OLGA/OLGIM low-risk gastritis: all 6 were diffuse-type gastric cancers (Figure 2). Two intestinal-type GCs coexisting with OLGA stage III were associated with OLGIM stage II gastritis (Figures 2 and 3).

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Distribution of preneoplastic/neoplastic lesions in the two considered staging-systems (i. e., operative link for gastritis assessment (OLGA)] and operative link on intestinal metaplasia assessment stages (OLGIM)]. ¹Diffuse-type (signet ring) gastric cancer. GC: Gastric cancer; LG-IEN: Low-grade intraepithelial neoplasia; HG-IEN: High-grade intraepithelial neoplasia.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 One of two gastric cancers developing in operative link for gastritis assessment stage III, but the operative link for gastritis assessment stage II, gastric mucosa. Representative images of 6 biopsy samples (3 from the mucosecreting/antral compartment, 1 from the lesion, and 2 from the oxyntic/corpus compartment) labeled according to site of origin (A: antral/angular; GC: gastric cancer; C: corpus), showing the percentages of atrophic/metaplastic lesions and the consequent operative link for gastritis assessment/operative link on intestinal metaplasia assessment stages.

Gastric cancer is still a health priority in Western Europe and it represents an epidemiological emergency in Eastern Europe, Central and Eastern Asia, and some South American regions[17-26].

Gastric mucosal atrophy is generally considered the “cancerization field” in which GC develops. Based on such a rationale, and incorporating the experience gained with the Sydney system[27], the OLGA staging system ranks the gastritis-associated cancer risk according to both the topography and the extent of gastric mucosa atrophy[2,4-6,26-28].

As regards topography, extensive biopsy sampling protocols (such as the one applied in the Houston experience) potentially increase the prognostic reliability of any staging system and they should be theoretically preferred. In line with the Sydney system[27], however, both OLGA and OLGIM systems require a (minimum) set of 5 biopsy samples for gastritis staging, which should strike a good compromise between the priority of obtaining a representative biopsy set and the operative limits of daily clinical practice[27].

The OLGIM proposal replaces the “global” atrophy score with a semiquantitative assessment of intestinal metaplasia (extent and site); according to its proposers, such a strategy should considerably increase the inter-observer agreement - an undeniable advantage[29,30].

In the present series of more than four thousand consecutive cases, 98% of stage III-IV gastritis were consistently staged by applying either OLGA or OLGIM. The finding that patients’ ages increase with higher stages further supports the clinico-biological plausibility of both systems.

It is worth noting that two intestinal-type GCs (both coexisting with OLGA stage III gastritis) were found associated with OLGIM-II gastritis (i.e., low-risk atrophic gastritis). In fact, by focusing on IM alone, OLGIM is less sensitive in identifying high-risk gastritis, and this may result in the down-staging of patients who should be offered follow-up[29,30]. Comparative studies involving non-GI (i.e., specialist) pathologists are needed to test which system (OLGA or OLGIM) provides more accurate results in relation to the time and effort spent on the histology assessment.

In his seminal work on gastric carcinogenesis, Pelayo Correa described mucosal atrophy as a cardinal step in the biological pathway that may eventually progress to gastric adenocarcinoma[31]. The current definition of gastric mucosal atrophy includes two different phenotypes: (1) loss (shrinkage or disappearance) of glands, which are replaced by fibrotic expansion of the lamina propria; and (2) metaplastic replacement of native glands by intestinalized and/or pseudopyloric glands (corpus antralization or SPEM). Focusing on IM alone excludes pseudopyloric metaplasia (i.e., SPEM) from the spectrum of atrophy, although it has recently been found increasingly important in gastric carcinogenesis (through transdifferentiation from mature chief cells following parietal cell loss)[32-34].

Lastly, considering IM alone carries the risk of us losing the correlation between gastric atrophy (as assessed by gastric serology, and PgI in particular) and its organic counterpart (resulting from the concurrence of the different phenotypes of gastric atrophy)[35-38].

In conclusion, gastritis staging effectively conveys an unequivocal message regarding the gastritis-associated cancer risk and may point to follow-up strategies tailored to a patient-specific clinico-pathological situation. This priority supports the inclusion of staging in gastritis histology reports and the demand for further efforts to improve the reproducibility of any staging criteria - bearing in mind that “easier” does not necessarily mean “better”!