First‐degree relatives of patients with early‐onset gastric carcinoma show even at young ages a high prevalence of advanced OLGA/OLGIM stages and dysplasia
Summary
Background
First‐degree relatives (FDRs) of early‐onset gastric carcinoma (EOGC) patients are at increased risk of cancer development. OLGA/OLGIM (Operative Link on Gastritis/Intestinal Metaplasia Assessment) classifications have been proposed for the identification of individuals at high risk of gastric cancer development.
Aim
To estimate the prevalence and severity of premalignant conditions and lesions in FDRs of EOGC patients.
Methods
A case–control study was conducted encompassing 103 FDRs of EOGC patients (cases) and 101 age‐ and gender‐matched controls, all submitted to upper GI endoscopy and OLGA and OLGIM used for staging as well as modified versions with exclusion of the biopsies from incisura angularis in the analysis.
Results
Helicobacter pylori infection was present in 82% of cases (P = 0.001). Atrophy was present in 70% of cases (OLGA stages I–IV). High‐risk stages (III–IV) were identified only in cases (19%) (P < 0.001). Dysplasia was diagnosed only in cases (n = 7, P = 0.007). The application of OLGIM, modified OLGA and modified OLGIM classifications led to downgrade of stages in comparison with the original OLGA classification (27%, 15% and 30% respectively). In all classification systems, dysplastic lesions clustered (86%) in high‐risk stages.
Conclusions
FDRs of EOGC patients have, even at young ages, a high prevalence of H. pylori infection, high‐risk OLGA and OLGIM stages and dysplasia. These patients should undergo accurate endoscopic observation with at least four biopsies in antrum and corpus to allow adequate staging and follow‐up of premalignant conditions and lesions scored in high‐risk stages, in accordance with international guidelines recently proposed.
Introduction
Despite its declining incidence in Western Countries, gastric cancer is the second most lethal neoplasm worldwide.1, 2 At least for Lauren's intestinal type of gastric adenocarcinoma,3 a cascade of associated lesions has been defined and known as ‘Correa's Cascade’, which describes a multistage progressive sequence of lesions (chronic atrophic gastritis, intestinal metaplasia (IM), dysplasia and invasive gastric carcinoma).4 The risk of cancer increases with the extension of these lesions.5-7 The most widely accepted system for diagnosing and grading preneoplasic lesions is the Sydney and the updated Sydney System.8 Although those classification systems brought some uniformity, they do not provide information about gastric cancer risk. The recently established OLGA staging system (Operative Link for Gastritis Assessment) aims to translate the histopathological data into a standardised report with information on the gastric condition (topography and extent of the atrophic changes), and subgrouping of patients by cancer risk using the Sydney‐Houston five biopsy protocol.9 The gastritis stage results from the combinations of the atrophy score (at the single biopsy level) with the atrophy topography (as obtained by gastric mapping). Another proposed classification system used IM as a phenotypic marker of atrophy (OLGIM – Operative Link on Gastric Intestinal Metaplasia Assessment).10 Recently, guidelines on follow‐up of gastric premalignant lesions and conditions of the stomach recommended four biopsies (two of the antrum and two of the corpus) for adequate staging these lesions, assuming that the biopsy of incisura angularis offers little advantage for defining the spread/extension of the lesions, which is the parameter that confers risk for gastric cancer.5, 6, 11
Environmental and genetic factors have a role in the aetiology of gastric cancer. Smoking, dietary habits, and Helicobacter pylori infection have been pointed out as important risk factors. H. pylori plays a pivotal role in this progression and has been classified as a type 1 carcinogen in 1994 by WHO.12 It is believed that the combination of a virulent microorganism in a genetically susceptible host is necessary for continued chronic inflammation and gastric carcinogenesis.13-15 The risk of stomach cancer is increased in subjects with family history of the disease, and epidemiological (case–control) studies reported odds ratios of 2–10, varying with the geographical region and ethnicity.16-23 The familial clustering of cancer may occur due to inherited genetic susceptibility, shared environmental or lifestyle factors, or a combination of these. Some studies found a higher prevalence of H. pylori infection, gastric atrophy and IM in first‐degree relatives of gastric cancer patients.24 Some patients develop gastric cancer before 45 years of age (early‐onset gastric cancer – EOGC) and it is believed that the clinico‐pathological profile of these patients is different from that of patients developing gastric cancer at a later age. It is believed that gastric carcinogenesis is accelerated in EOGC due to genetic susceptibility factors.25
The aim of this work was to estimate the prevalence and severity of premalignant conditions and lesions in FDRs of EOGC patients using OLGA and OLGIM systems (and two modified versions of these classifications).
Methods
Type of study and selection of participants
A case–control study that included first‐degree relatives of early onset gastric carcinoma patients (i.e. diagnosed before 45 years) was performed. Clinical records of all gastric cancer patients treated at Centro Hospitalar do Porto, between 2005 and 2010 were examined and age of presentation and data on histopathology of noncardia gastric carcinoma were retrieved. After exclusion of families with other type of cancers pointing to the possibility of hereditary nonpolyposis colorectal cancer (fulfilling Amesterdam II criteria26), 40 index patients were selected. First‐degree relatives (siblings and children) were invited to participate in the study.
A maximum of 160 cases (first‐degree relatives of the 40 index patients) were considered eligible and 103 individuals accepted to be enrolled in the study. In index cases, gastric carcinoma was classified as diffuse (n = 20), intestinal (n = 15) and mixed (n = 5) according to Lauren classification.3 The mean age of diagnosis of gastric cancer index patients was 41 years (range 31–44 years).
Controls encompassed spouses and neighbours of the cases (n = 50) and the remaining controls (matched for age, gender and smoking habits) were recruited from dyspeptic patients submitted to upper endoscopy for investigation of dyspepsia (n = 51). Validity of control selection has been described elsewhere.27
Cases and controls younger than 18 years or older than 70 years were excluded. Additional exclusion criteria were serious or debilitating renal, liver, blood or mental disorders and previous gastric surgery. Confidentiality and clinical assistance were guaranteed whenever a gastric pathological condition was diagnosed.
Procedures
The study was approved by the ethical committee of Centro Hospitalar do Porto and informed consent was obtained. Patients were asked to answer a questionnaire regarding demographic data (age, gender, area of residence and smoking habits).
All cases and controls underwent upper endoscopy and biopsies were obtained for histopathological study and H. pylori culture.
Endoscopic procedure
A conventional upper GI endoscopy was performed by the same Gastroenterologist (R.M‐P.) by using a High‐Definition forward‐viewing videogastroscope (GIF‐Q180; Olympus Optical Co., Ltd, Tokyo, Japan). All examinations were performed in an outpatient basis without sedation. Gastric biopsies were taken using the updated Sydney system: one from incisura, two from the antrum (3 cm from the pylorus, at greater and lesser curvatures) and two from the corpus (one from lesser curvature, 4 cm proximal to the incisura, and one from the middle of the greater curvature). An additional antral biopsy was taken for H. pylori culture. Targeted biopsies were taken in case of identifiable mucosal lesions. No complications occurred as a consequence of endoscopy and biopsy procedures.
Histopathological evaluation
Gastric biopsies were fixed in 10% buffered formalin and embedded in paraffin. Sections obtained from paraffin blocks were stained with haematoxylin and eosin (H&E) and modified Giemsa. Histological evaluation was performed by two experienced pathologists (C.L. and X.W.) who were blinded to the patient data and endoscopic findings. All histological sections were assessed using the updated Sydney classification system and the following parameters were scored: H. pylori density, acute inflammation (neutrophil infiltration), chronic inflammation (mononuclear infiltration), gastric atrophy and IM. These items were scored as 0 (absent),1 (mild), 2 (moderate) and 3 (marked). The WHO classification was used for the assessment of dysplasia.28 On the basis of atrophy topography (oxyntic and antral/angular), the gastritis stage was assessed according to the international OLGA proposal29 (Figure 1). For OLGIM system, IM was used as the phenotypic marker of gastric mucosal atrophy.10 We also considered a modified OLGA and a modified OLGIM grading by using the same principles of OLGA and OLGIM, respectively, but with suppression of the incisura angularis biopsy, thus encompassing two biopsies of the antrum and two biopsies of the corpus (for assessment of atrophy in these localizations). High‐risk stages were defined as stages III–IV for all classifications.
Helicobacter pylori infection status
Gastric biopsies were evaluated for the presence of H. pylori using sections stained with modified Giemsa. Additionally, culture from fresh biopsies was performed. H. pylori infection was considered as present whenever the microorganisms were identified by histopathology and/or culture.
Statistical analysis
spss 18.0 (SPSS Inc., Chicago, IL, USA) was used for data analysis. Parametric and/or nonparametric tests both for continuous variables or ordinal or dichotomic variables were used when applicable. A 5% significance level was considered.
Results
The mean age of the 103 first‐degree relatives of EOGC (cases) was 41 years (range 31–44 years) and 40% were men. Controls were matched for gender and age.
Globally, chronic atrophic gastritis (CAG) was diagnosed in 70% of cases and in 32% of the controls (P < 0.001). H. pylori infection was detected in 82% of the cases compared to 65% in controls (P = 0.001). Overall, six active duodenal ulcers were identified, two in cases and four in controls; no gastric ulcers were detected. All duodenal ulcers as well all dysplastic lesions were detected in H. pylori‐positive patients. Seven cases of dysplastic lesions were detected in cases and none in controls (P = 0.007), encompassing one high‐grade dysplasia and six cases of low‐grade dysplasia. Five dysplasia lesions were detected in antrum and two in corpus. The mean age of cases with dysplasia was 53 years.
Gastritis staging
OLGA staging
Tables 1 and 2 show OLGA gastritis stage distribution in cases and controls. The results show: stage 0:30% vs. 68%; stage I: 40% vs. 30%, stage II: 12% vs. 2%, in cases and controls respectively. Stages III and IV (high‐grade stages) were identified only among cases (10% and 9% respectively) (P < 0.001). All peptic ulcers clustered in low‐stage gastritis. Dysplasia was only detected in cases, 86% of them were included in high‐risk stages (III–IV) (P < 0.005) (Table 2), 29% in stage III and 57% in stage IV. Only one case of dysplasia occurred in stage II (14%). Dysplasia was not identified in stages 0 and I.
| Cases | Controls |
P
aa
95% CI.
|
|
|---|---|---|---|
| n = 103 | n = 101 | ||
| H. pylori positive (%) | 82 | 59 | 0.001 |
| Histological phenotype | |||
| CAG (%) | 70 | 32 | <0.001 |
| OLGA stage (%) | |||
| 0 | 30 | 68 | <0.001 |
| I | 40 | 30 | |
| II | 12 | 2 | |
| III | 10 | 0 | |
| IV | 9 | 0 | |
| Dysplasia (%) | 7 | 0 | 0.007 |
- a 95% CI.
| Atrophy Score | Corpus | |||
|---|---|---|---|---|
| No atrophy (0) | Mild atrophy (1) | Moderate atrophy (2) | Severe atrophy (3) | |
|
Cases n = 103 |
||||
| Antrum | ||||
| No atrophy (0) | 31 | 5 | 0 | 0 |
| Mild atrophy (1) | 25 | 11 | 3 | 0 |
| Moderate atrophy (2) | 4 | 5 | 8 | 0 |
| Severe atrophy (3) | 0 | 2 | 6 | 3 |
|
Dysplasia (cases) n = 7 |
||||
| Antrum | ||||
| No atrophy (0) | 0 | 0 | 0 | 0 |
| Mild atrophy (1) | 0 | 0 | 0 | 0 |
| Moderate atrophy (2) | 0 | 1 | 1 | 0 |
| Severe atrophy (3) | 0 | 1 | 2 | 2 |
|
Controls n = 101 |
||||
| Antrum | ||||
| No atrophy (0) | 67 | 0 | 0 | 0 |
| Mild atrophy (1) | 25 | 5 | 0 | 0 |
| Moderate atrophy (2) | 3 | 1 | 0 | 0 |
| Severe atrophy (3) | 0 | 0 | 0 | 0 |
- OLGA, operative link for gastritis assessment.
- Grey shades represents high‐risk stages (III–IV).
OLGIM staging vs. OLGA staging
Seventy‐one per cent of the cases were classified as stages I–IV according to the OLGA system: stage I (n = 41); stage II (n = 12); stage III (n = 10), stage IV (n = 9); 44% of cases were classified as stages I–IV according to OLGIM staging: stage I (n = 21); stage II (n = 7); stage III (n = 9); stage IV (n = 7) (Tables 2 and 3). Taking into account the high‐risk stages (III–IV), 19% and 16% of the cases were scored as stages III–IV in OLGA and OLGIM systems respectively. For dysplastic lesions, the distribution in high‐risk stages was the same in OLGA and OLGIM classifications (86%; n = 6), showing a strong association between the presence of dysplasia and high‐risk stages. The application of OLGIM system led to a downgrade of stage in 27% of cases. Within the high‐risk stages, the downgrading of stage was observed only in 3% of the cases, with maintenance of the clustering of dysplastic lesions (86%) in advanced stages. Regarding the whole series (cases and controls, n = 204), the global downgrade of stages was 19% when comparing OLGIM with OLGA classification (Table 4).
| Atrophy Score | Corpus | |||
|---|---|---|---|---|
| No atrophy (0) | Mild atrophy (1) | Moderate atrophy (2) | Severe atrophy (3) | |
|
Cases n = 103 |
||||
| Antrum | ||||
| No atrophy (0) | 59 | 2 | 0 | 0 |
| Mild atrophy (1) | 14 | 5 | 2 | 0 |
| Moderate atrophy (2) | 4 | 1 | 7 | 0 |
| Severe atrophy (3) | 0 | 2 | 5 | 2 |
|
Dysplasia (cases) n = 7 |
||||
| Antrum | ||||
| No atrophy (0) | 0 | 0 | 0 | 0 |
| Mild atrophy (1) | 1 | 0 | 0 | 0 |
| Moderate atrophy (2) | 0 | 0 | 2 | 0 |
| Severe atrophy (3) | 0 | 1 | 1 | 2 |
|
Controls n = 101 |
||||
| Antrum | ||||
| No atrophy (0) | 76 | 0 | 0 | 0 |
| Mild atrophy (1) | 22 | 2 | 0 | 0 |
| Moderate atrophy (2) | 1 | 0 | 0 | 0 |
| Severe atrophy (3) | 0 | 0 | 0 | 0 |
- OLGIM, operative link on gastric intestinal metaplasia assessment.
- Grey shades represents high‐risk stages (III–IV).
| Stage | Classification system | ||||||
|---|---|---|---|---|---|---|---|
| OLGA (n) | Dgde% (ModOLGA vs. OLGA) | ModOLGA (n) | Dgde% (OLGIM vs. OLGA) | OLGIM (n) | Dgde% (ModOLGIM vs. OLGA) | ModOLGIM (n) | |
|
Cases n = 103 |
|||||||
| 0 | 31 | – | 44 | – | 59 | – | 62 |
| I | 41 | 9% | 32 | 19% | 21 | 21% | 19 |
| II | 12 | 1% | 13 | 5% | 7 | 4% | 8 |
| III | 10 | 3% | 7 | 1% | 9 | 3% | 7 |
| IV | 9 | 2% | 7 | 2% | 7 | 2% | 7 |
| Dgde% (Total) | 15% | 27% | 30% | ||||
|
Dysplasia n = 7 |
|||||||
| 0 | 0 | 0 | 0 | 0 | |||
| I | 0 | 0 | 1 | 1 | |||
| II | 1 | 1 | 0 | 0 | |||
| III | 2 | 3 | 3 | 3 | |||
| IV | 4 | 3 | 3 | 3 | |||
|
Controls n = 101 |
|||||||
| 0 | 67 | – | 70 | – | 76 | – | 80 |
| I | 30 | 2% | 28 | 6% | 24 | 10% | 20 |
| II | 4 | 1% | 3 | 3% | 1 | 3% | 1 |
| III | 0 | – | 0 | – | 0 | – | 0 |
| IV | 0 | – | 0 | – | 0 | – | 0 |
| Dgde% (Total) | 3% | 9% | 13% | ||||
|
Total n = 204 |
|||||||
| 0 | 98 | – | 114 | – | 135 | – | 142 |
| I | 71 | 6% | 60 | 13% | 45 | 16% | 39 |
| II | 16 | – | 16 | 4% | 8 | 3% | 9 |
| III | 10 | 1% | 7 | 0.5% | 9 | 1% | 7 |
| IV | 9 | 1% | 7 | 1% | 7 | 1% | 7 |
| Dgde% (Total) | 8% | 19% | 22% | ||||
- Dgde, downgrade of stage compared with OLGA classification system; OLGA, operative link for gastritis assessment; OLGIM, operative link on gastric intestinal metaplasia assessment.
- Grey shades represents high‐risk stages (III–IV).
Modified OLGA and OLGIM staging (excluding incisura angularis biopsy) vs. OLGA staging
The comparison of modified OLGA and modified OLGIM staging systems (which excludes the biopsy from incisura angularis) with OLGA staging showed also a general downgrade of stages for the cases (Table 4). As mentioned above, 70% of the cases were classified as stages I–IV according to the OLGA staging: stage I (n = 41); stage II (n = 12); stage III (n = 10); stage IV (n = 9); 55% of cases were classified as stages I–IV according to modified OLGA system: stage I (n = 32); stage II (n = 13); stage III (n = 7); stage IV (n = 7); 40% of cases were classified as stages I–IV according to modified OLGIM system: stage I (n = 19); stage II (n = 8); stage III (n = 7); stage IV (n = 7). The downgrade of high‐risk stages (III–IV) was 5% for both, modified OLGA and modified OLGIM, when compared with OLGA system. The distribution of dysplasia in high‐risk stages was the same for the three staging systems (86% clustered in stages III–IV). Taking into account the whole series (cases and controls, n = 204), there was a global downgrade of stages of 8% (modified OLGA vs. OLGA) and 22% (modified OLGIM vs. OLGA) (Table 4).
Discussion
To the best of our knowledge, this study reports for the first time the changes in gastric mucosa of FDRs of EOGC patients. The results show that there is a high prevalence of H. pylori infection (81% vs. 59% in controls) and premalignant conditions/lesions, with 70% of cases displaying CAG. H. pylori infection and CAG are more prevalent in this sample of relatives of EOGC patients than in other series encompassing relatives of gastric cancer patients, regardless of the age of onset.24, 30-37 In those studies, prevalence of GAC ranges from 11% to 30%. We observed also a high percentage of dysplasia (7%) in relatives of EOGC patients (mean age of 53 years). These results support the hypothesis of an accelerated carcinogenic process in this population. Genetic susceptibility is considered to play a major role in addition to environmental factors, H. pylori being determinant.
The classification of gastritis using OLGA system aims to translate the histopathological data into a standardised report (with information on topography and extent of the atrophic changes) and sub grouping patients by cancer risk (those included in stages III and IV harbouring the highest risk for gastric cancer development, therefore being eligible for follow‐up),9 as suggested in two cross‐sectional studies.38, 39 Recently, a 12‐year follow‐up study gave additional support to the initial results40 and showed also a relationship of stages III–IV with lower pepsinogen I/II ratio, a marker of gastric extensive atrophy.40, 41 IM subtyping was also shown to be incorporated in high‐risk OLGA stages (III and IV) with sulfomucin‐type IM (type III) being associated with an increased risk of gastric cancer.42 In our study, the first case–control study applying OLGA staging to a high‐risk population, 19% of cases (FDRs of EOGC patients) clustered in stages III–IV compared to 0% of the controls (P < 0.001), showing that there was a high number of patients with extensive atrophy, thus at higher risk for intestinal type gastric adenocarcinoma and eligible for follow‐up. All peptic ulcers were included in low‐risk stages and 86% of dysplastic lesions clustered in stages III–IV (P < 0.005).
It has been recognised that a potential drawback of OLGA staging is the fact that it relies on the severity and extent of gastric atrophy, a parameter for which there is low inter‐observer agreement.8, 43, 44 Capelle et al. proposed a classification system (OLGIM) that uses only one of the phenotypic expressions of atrophy (IM), achieving a higher inter‐observer agreement (K = 0.9 for IM and K = 0.6 for atrophic gastritis) and a downgrade of stages leading to a smaller population suitable for surveillance10. 10 Recently, Rugge et al. conducted a retrospective, nonstandardised study, encompassing 4552 consecutive biopsy sets, aiming at the comparison of OLGA and OLGIM classification systems.45 Inter‐observer agreement was found to be higher for OLGIM than for OLGA system and the use of the former resulted in a 6.6% downgrade of stages in comparison with OLGA classification. In both staging systems, a significant association was observed between stages III–IV and gastric neoplasia.45
In our study, 44% of cases were scored as stages I–IV according to OLGIM with a 27% downgrade in comparison with OLGA. In stages III–IV, there was a 3% downgrade of staging, without interfering with the clustering of dysplasia (86%) in high‐risk stages, which is in keeping with the results obtained by the proposers of the OLGIM system.10
The value of the diagnostic yield of biopsy of incisura angularis has been addressed in several studies. Despite being considered the earliest location of the onset of atrophy/metaplastic transformation,38, 46 at least two studies concluded that few histological findings were detectable only in biopsies of the incisura angularis.47, 48 These results suggest that more important than the site where the atrophic transformation begun is its extension, the latter being the parameter that confers gastric cancer risk.5-7 Guidelines recently proposed for the follow‐up of gastric premalignant conditions and lesions excluded the biopsy of the incisura angularis for staging, keeping two biopsies of the antrum and other two biopsies of the corpus.11 In the group of FDRs of EOGC patients, using modified OLGA and modified OLGIM staging systems (with exclusion of the biopsy of incisura angularis), we observed, respectively, 15% and 30% downgrade of staging in comparison with the original OLGA system. Equal and less pronounced downgrade of staging was observed in high‐risk (5%) for both for both modified staging systems in comparison with OLGA system. A relevant fact is that for all staging systems, advanced lesions (e.g. dysplasia) cluster in high‐risk stages.
The present study has a cross‐sectional design focusing on a sample of Portuguese population, which is a high‐risk population for gastric cancer. Furthermore, endoscopy was performed in asymptomatic relatives of gastric cancer patients achieving a participation of 64% of initially eligible cases, which may result in selection bias. Future large, multicentre, prospective studies are needed to confirm our results.
Summing up, the results of this study show that in FDRs of EOGC patients, there is a high prevalence of H. pylori infection and CAG and an important number of these individuals were classified in high‐risk stages (III–IV) of OLGA and OLGIM classifications (and their modified versions with exclusion of the biopsy of incisura angularis). OLGIM and modified staging systems led to downgrade of stages in comparison with the original OLGA system, thus identifying a lower number of individuals eligible for follow‐up. Importantly, dysplastic lesions cluster in high‐risk stages, regardless of the staging system under analysis. We estimate that OLGIM will be easier, as less time consuming and equally accurate. Nevertheless, further reliability and validation studies are needed confirm our results
On the basis of these results, we recommend that FDRs of EOGC patients should be submitted to accurate endoscopic observation with at least four biopsies in antrum and corpus to allow adequate staging and follow‐up of premalignant conditions and lesions scored in high‐risk stages, in accordance with international guidelines recently proposed.11
Acknowledgements
The authors thank Prof. Leonor David for helpful suggestions and discussion and to Dr. Sollari Allegro for the support. Declaration of personal interests: None. Declaration of funding interests: This study was funded by "Fundação para a Ciência e Tecnologia" (SFRH/SINTD/60034/2009), (PTDC/SAU‐SAP/120024/2010), (SFRH/BD/45841/2008) and Hospital Santo António.
