The association between gastric endoscopic findings and histologic premalignant lesions in the Iranian rural population.

Atrophic gastritis, intestinal metaplasia, and gastric dysplasia are histologic premalignant lesions (PMLs). Correlation between the gastric endoscopic findings and histologic PMLs is not clear. This study was designed to determine the possible association of endoscopic findings and histologic PMLs.Over 28 months gastric endoscopic findings of consecutive rural patients with dyspepsia were categorized into 3 groups: 1-normal, 2-ulcerative with or without concurrent abnormality, 3-abnormal non-ulcerative. Biopsies of antrum and body were taken from all included patients and examined for the presence of histologic PMLs. Any mucosal abnormality was also biopsied.From 7340 evaluated patients, an overall of 1973 patients were included. 55.7% of patients were in group 1; 3.8% in group 2 and 40.5% in group 3. A within sex analysis showed that the majority of male patients were in PMLs subgroup (P < 0.001) likewise in groups 2 and 3 (P < 0.001). The prevalence of histologic PMLs in groups 2 and 3 was significantly higher than group 1 (P < 0:001) but the difference was not significant between groups 2 and 3 (P = 0.484). Mean (±SD) age of patient with PMLs was 50.25 ± 17.71 whereas in patients without PMLs was 41.16 ± 16.48 (P < 0.001).This study has showed that abnormal gastric endoscopic findings, male sex and increased age can be considered as risk factors of the formation of histologic PMLs. Until further investigations we propose that any abnormality on gastric mucosa (ulcerative or non-ulcerative) could be biopsied for the evaluation of probable histologic PMLs especially in old men.

INTRODUCTION

Gastric cancer is the fifth prevalent cancer and the third cause of cancer-related mortality worldwide.[1] Atrophic gastritis (AG), intestinal metaplasia (IM), and gastric dysplasia (GD) are histologic premalignant lesions (PMLs) which are considered as the multistep cascade precursors of gastric cancer development.[2-5]

The presence and extension of PMLs in the stomach are frequently diagnosed by endoscopic or histologic examinations.[6,7]

Dyspepsia as a common symptom is used for the characterization of epigastric pain or discomfort. Although the functional disorder is a major cause of dyspepsia, organic conditions such as ulcer and gastric cancer can also be present in dyspepsia.[8,9]

Endoscopy is generally used for the diagnosis of the possible causes of dyspepsia. In general practice, different appearances can be found during endoscopy; however, there is no consensus about the association of gastric endoscopic findings and histopathological conditions especially PMLs on which there are only a few published researches, which have described this controversial correlations.[10-25]

Although poor correlation between endoscopic findings and histologic changes was detected in many studies,[22-24,26-30] good correlation was reported only in the severe types of gastritis or normal endoscopy.[25,26,31]

Up to now there is no clear correlation between endoscopic findings of stomach and histologic PMLs, for example, AG, IM, and GD according to few published researches, so we designed this study to investigate this possible association. In this research 3 groups of endoscopic findings were comparatively analyzed for the existence of histologic PMLs in a population of adult patients with dyspepsia.

MATERIAL AND METHODS

Population

In order to evaluate the prevalence and association between gastric endoscopic finding and histologic PMLs, consecutive Iranian rural patients with dyspepsia were enrolled. Cases with the following conditions were excluded: (a) previous Helicobacter pylori eradication; (b) history of gastric or esophageal surgery; (c) positive history of recent treatment (up to 2 weeks prior to inclusion) with H2-blockers, proton pump Inhibitors, and NSAIDs; (d) evidence of malignancy in histopathological examination; (e) patients with poor cooperation.

Study Design

This study was carried out between November 2011 and March 2014. The study protocol was approved by the Hospital Ethics Committee. Written informed consent was obtained from each patient or his or her legal guardian in accordance with the Helsinki Declaration.

Dyspepsia was defined as the epigastric pain or discomfort.[8,9]

The high resolution white light endoscopic procedure (PENTAX Video Processor: 300W Xenon lamp-PENTAX Video Gastroscope: EG-290Kp with insertion tube 9.8 mm) was performed by an expert gastroenterologist for all included patients.

The gastric endoscopic findings were categorized into 3 groups:

1—normal, 2—ulcerative with or without concurrent abnormality, 3—abnormal non-ulcerative (any evidence of mucosal abnormality except ulcerative lesion).

Two biopsies of antrum and 2 biopsies of body were taken from all of the enrolled cases with normal or abnormal endoscopy. Biopsy of the abnormal lesions was also performed for any mucosal abnormality. The biopsy samples were fixed in 10% formalin and examined by 2 expert pathologists for the presence of histologic PMLs, for example, atrophic gastritis, IM, and dysplasia.

Atrophic gastritis was defined as a severe decrease in typical gastric glands with the presence of inflammation.[32,33] IM was defined as the replacement of the glandular gastric mucosa with epithelium similar to the intestinal type cells.[32,34] Gastric dysplasia was defined as the cellular pleomorphism, nuclear hyperchromatism, and increase in nuclear/cytoplasmic ratio, decreased cytoplasmic mucin, increased mitotic activity, and glandular disarray.[21]

Statistical Analysis

Comparative analysis of the groups was performed by using chi square test for categorical variables and by t test for continuous variables. Odds of histopathological features were analyzed by using logistic regression. The 95% confidence interval for key proportions was calculated using the exact binomial distribution. A P-value of 0.05 or less was considered statistically significant. Statistical analysis was performed with SPSS 15.0.

RESULTS

From 7340 evaluated patients, an overall of 1973 patients (1288 females and 685 males) were enrolled. The mean age (SD) of patients was 42.61 (±17.024) years old (Table 1).

TABLE 1

The Characteristics of the Patients With and Without Histologic Premalignant Lesions (PMLs)

On endoscopic examination, 1098 (55.7%) patients showed normal endoscopy (group 1), 76 patients (3.8%) had ulcer with or without concurrent abnormality (group 2) and 799 patients (40.5%) had abnormal non-ulcerative endoscopic findings (group 3) (Table 2).

TABLE 2

The Characteristics of the Gastric Endoscopic Findings

Overall the majority of enrolled patients were female. On histologic examination, 277 patients had an evidence of histologic PMLs while 1696 patients showed no evidence of histologic PMLs. A within sex analysis showed that the majority of male patients were in PMLs subgroup (P < 0.001; Chi-square test). (Table 1)

Moreover; 59.2% of female patients showed normal endoscopy (group 1) in contrast of 49.0% of male patients in this group. The within sex analysis in groups 2 (P = 0.003) and 3 (P < 0.001) in comparison with group 1 showed that the predominance of female gender had normal endoscopy, but there was not any significant difference between groups 2 and 3 (P = 0.162).

Abnormal non-ulcerative endoscopic findings of our cases included erosion, erythema, nodularity, atrophic mucosa, and polypoid lesion.

Age of patients with histologic PMLs was significantly higher than patients without histologic PMLs (P < 0.001) (Table 1).

Age of patients with abnormal endoscopy (groups 2 and 3) was significantly higher than normal endoscopy (P < 0.001; multiple comparisons based on Tukey test).

The prevalence of histologic PMLs in patients with ulcer (group 2) was significantly higher than that of normal endoscopy group (group 1). The prevalence of histologic PMLs in patients with abnormal non-ulcerative endoscopic findings (group 3) was also significantly higher than normal endoscopy group (group 1). Furthermore, the prevalence of histologic PMLs showed no significant difference between groups 2 and 3 (P = 0.484) (Table 3).

TABLE 3

Association Between Gastric Endoscopic Findings and Histologic Premalignant Lesions

DISCUSSION

There are a few published researches about the correlation between endoscopic findings and histologic PMLs. As far as we know this study is the first report of Iranian adult population with dyspepsia in which a range of histologic PMLs were comparatively analyzed between the 3 groups of gastric endoscopic findings.

Association between GD as a histologic PML and gastric endoscopic findings is not also clear according to previous studies. A variety of endoscopic findings including ulceration, atrophic mucosa, polyps, erosions, plaques, and scars have been associated with GD. GD can also be associated with a normal endoscopy.[16-18,21] Lansdown et al[17] evaluated patients with GD. On review, only 20 of the 40 patients had true dysplasia. The endoscopic findings of these cases included ulcer (9/20), raised tumor or polyp (6/20), plaque (1/20), atrophic mucosa (2/20), irregular mucosa at a gastroenterostomy (1/20), deformity of the pyloric canal (1/20). In another study, Aste et al showed that 29 of 694 patients with endoscopic localized gastric lesions (ulcers, erosions, enlarged or irregular folds, tumors and polyps) had GD but only 1 of 123 patients with normal endoscopy had GD. They concluded that GD is significantly associated with prominent or depressed lesions on endoscopy.[35]

Although some endoscopic criteria for AG have been described, a few reports have showed that the diagnosis of AG by conventional endoscopy often correlates poorly with histology.[7,36] Lin et al in a prospective study had evaluated the accuracy of the diagnosis of IM by endoscopy and the correlation of endoscopic diagnosis with histology. They showed that the specificity, sensitivity, and accuracy of the endoscopic diagnosis was 68.1%, 75%, and 71.3%, respectively.[28] Association between nodular gastritis as an endoscopic finding and histologic premalignant or malignant lesions is not also clear.[10-15]

Some studies have showed that there is a poor correlation between endoscopic findings and histological diagnosis of gastritis.[22,24,37,38] Kaur et al[39] in a study for evaluation of the correlation between endoscopic findings and histological gastritis showed that there was a poor correlation between them. They concluded that endoscopic finding is an unreliable predictor of histological gastritis. A study by Fung et al[24] showed that in dyspeptic patients endoscopic diagnosis is relatively inaccurate in specific types of gastritis. They showed that among 33 dyspeptic patients endoscopically diagnosed gastritis, histological confirmation was detected in 3/9, 10/14, and 0/6 cases of chronic atrophic gastritis, chronic (superficial) gastritis, and acute gastritis, respectively. A study by Redéen et al[27] on 488 adult individuals selected from a general population showed that except for the absence of visible vessels and rugae in the gastric corpus, endoscopic findings had very limited value in the evaluation of histological gastritis. Their endoscopic findings were erythema, erosions, presence of visible vessels, and absence of rugae in the gastric corpus. Calabrese et al[38] in a prospective study for the evaluation of the correlation of endoscopic findings with histologic changes and H pylori infection showed that the correlation between endoscopic findings and histological diagnosis of gastritis is poor. They concluded that biopsies are mandatory in all patients. A study by Jönsson et al[23] showed that in 210 dyspeptic patients endoscopic diagnosis correlates significantly with the histologic changes in the bulb of duodenum but not in the stomach.

Abnormal endoscopic findings were also seen in an asymptomatic population. In a prospective study by Akdamar et al[40] on 355 asymptomatic adult individuals showed that 86 (24%) of them had abnormal endoscopic findings in the stomach. In a case control study Toukan et al[25] showed that normal endoscopy was more likely associated with a histologic normal neutrophil count. However there was no correlation between abnormal endoscopic finding and increased histologic normal neutrophil count.

In our report the prevalence of histologic PMLs in patients with ulcer was significantly higher than normal endoscopy. Histologic PMLs in patients with abnormal non-ulcerative endoscopy were also significantly higher than that of normal endoscopy (P = 0.000) (Table 3). However, the prevalence of histologic PMLs was not significantly different between endoscopic ulcerative group and abnormal non-ulcerative group (P = 0.484) (Table 3). These results have showed that abnormal endoscopic findings (ulcerative or non-ulcerative) can be considered as risk factors for the formation of histologic PMLs.

In this study, abnormal endoscopic findings (ulcerative and/or non-ulcerative) in comparison with normal endoscopy were significantly higher in male sex. Patients’ sex with histologic PMLs rather than cases without histologic PMLs was also significantly in favour of male gender. Similar to some studies, our result showed that male gender can be considered as a risk factor for the formation of histologic PMLs and/or abnormal gastric endoscopic findings.[21]

Age of our patients with histologic PMLs was significantly higher than patients without histologic PMLs (P < 0.001) (Table 1) that was similar to many previous studies.[21,41-43] Age of patients with abnormal endoscopy (groups 2 and 3) was also significantly higher than those with normal endoscopy (P < 0.001). These results have showed that increased age can be considered as a risk factor for the formation of histologic PMLs and/or abnormal gastric endoscopic findings as well.

Among patients with normal gastric endoscopy, 6.8% had histologic PMLs. This result has showed that histologic PMLs can also be seen in cases with normal gastric mucosa. Similar result was seen by Di Gregorio et al.[18] They evaluated 99 patients with the diagnosis of GD, out of which 7 had normal endoscopy. Abnormal endoscopic findings of their cases included ulcer, polypoid, erosion, atrophy, and scar.

Our study had some limitations. First, only white light endoscopy without new advanced endoscopic procedure such as narrow band imaging or chromoendoscopy was used. Second, biopsy of fundus of stomach with endoscopic normal mucosa was not taken. Third, only symptomatic rural patients in a single-center study were evaluated.

Based on our results it is recommended that more studies should be carried out to clarify the different types of abnormal gastric endoscopic findings as the possible causes of histologic PMLs. Until further investigations we propose that any abnormality on gastric mucosa (ulcerative or non-ulcerative) could be biopsied in all patients for the evaluation of possible histologic PMLs in addition to detection of helicobacter infection. Furthermore, although histologic PMLs were significantly higher in patients with abnormal endoscopy, normal mucosa in endoscopy cannot exclude the existence of histologic PMLs especially in a high risk population. Therefore until future researches we propose that in dyspeptic old age men, even with normal gastric mucosa in conventional endoscopy, random sampling can be considered for the detection of possible histologic PMLs.