Predictability of Gastric Intestinal Metaplasia by Mottled Patchy Erythema Seen on Endoscopy

Background Intestinal metaplasia (IM) is regarded as a premalignant lesion. However, endoscopic diagnosis of IM has been considered difficult. Using endoscopy, we found a unique pattern of erythema, “Mottled Patchy Erythema (MPE),” which includes severe IM. Helicobacter pylori (Hp) infection itself can cause erythema, which reflects histologic changes in the gastric mucosa. Therefore we enrolled Hp eradication patients to validate the relation between MPE and pathologic findings. Methods We enrolled patients with chronic gastritis who underwent successful Hp eradication at least 6 months before the study. We defined MPE as multiple flat or depressed erythematous lesions. When encountering MPE on endoscopy, we performed biopsy on both the MPE site and non-MPE site. The non-MPE site was defined as an adjacent mucosa located within 3 cm of the MPE site. All biopsy specimens were evaluated immunohistochemically for IM subtype using MUC2, MUC5AC, MUC6, CD10, and CDX2 stains. The degree of IM was defined according to the Updated Sydney System. The diagnostic accuracy of the MPE findings for pathologic IM was calculated. The relation between MPE and IM subtype was also assessed. Results A total of 102 patients were selected for the study. Of these, 55 (54%) patients had MPE. Biopsy specimens were taken from the MPE sites and non-MPE sites from these 55 patients. The IM percentages and median scores of IM were both significantly higher at the MPE sites (P < 0.001) than at the non-MPE sites. The sensitivity and specificity for MPE in the detection of histologic IM were 72.7% and 84.1%, respectively. No significant associations were observed in the expression of MUC2, MUC5AC, MUC6, CD10, and CDX2 between the MPE sites and non-MPE sites. There were no significant differences in the ratios (complete/incomplete) of IM subtypes between the two groups. Conclusions MPE is a useful endoscopic finding to detect histologic IM without the use of chromoendoscopy and magnifying endoscopy. However, the IM subtype is difficult to identify. In the era of Hp eradication, MPE has the potential to become a predictive finding for the risk of gastric cancer.


Introduction
It is believed that the development of gastric cancer involves a multi-step process, including Helicobacter pylori (Hp) infection, chronic gastritis, glandular atrophy, intestinal metaplasia (IM), and fi nally dysplasia [1]. IM and gastric atrophy are considered together to be risk factors for the development of intestinal-type gastric cancer and are regarded as premalignant lesions [2,3]. Gastric atrophy can be recognized by endoscopy and correlates with histologic evaluation [4,5]. However, the diagnosis of IM by using standard white light endoscopy has been considered to be diffi cult due to IM lacking distinction in color and its presence as multiple fl at lesions [6,7]. on endoscopy could be observed even after Hp eradication [8]. We describe this fi nding as "Mottled Patchy Erythema (MPE)." MPE can be recognized as multiple fl at or slightly depressed erythematous lesions under standard white light endoscopy; pathologically, it includes severe IM [8]. Hp infection itself can cause erythema, seen on endoscopy, which refl ects histologic changes such as infi ltration of infl ammatory cells and edema [9]. Therefore, we enrolled patients who underwent Hp eradication in order to validate the relation between MPE and pathologic fi ndings.

Patient selection
Patients with chronic gastritis who underwent successful Hp eradication at least 6 months prior to the study were prospectively enrolled for the study at the National Center for Global Health and Medicine (NCGM) between January 2008 and December 2008. Exclusion criteria included the use of nonsteroidal anti-infl ammatory drugs (NSAIDs), antacids, and anti-thrombotic drugs during the 4 weeks before endoscopy. We also excluded patients with a history of gastric surgery, hemorrhagic disease, liver cirrhosis, renal failure, heart failure, and early or advanced gastric cancer. Written informed consent was obtained from participants in accordance with the Declaration of Helsinki and its subsequent revision. The study protocol was approved by the Ethics Committee of the NCGM.

Helicobacter pylori eradication
Patients with chronic gastritis and peptic ulcer disease induced by Hp infection underwent eradication therapy. Patients were treated with a 7-day regimen consisting of amoxicillin, clarithromycin and a proton pump inhibitor (PPI) twice daily, which was the standard fi rst-line regimen approved in Japan. If eradication was not successful, a second regimen consisting of amoxicillin, metronidazole, and PPI was administered. Eradication was confi rmed by negative histologic examination of the gastric biopsies, together with a negative 13C-urea breath test (13C-UBT) 2 to 3 months after the completion of eradication therapy. When all of the tests were negative, a patient was defi ned as negative for Hp infection.

Endoscopic assessment
At least 6 months after the eradication of Hp, patients underwent endoscopic examination. We used a high resolution endoscope without magnifi cation (Olympus videoscope, model GIF-H260) to observe the presence of MPE in the gastric mucosa. We defi ned MPE as multiple fl at or slightly depressed erythematous lesions that were distinguishable from congested mucosa, hemorrhage, angioectasia, spotty erythema, and linear erythema (Fig. 1A, B). We also distinguished between MPE and reddish mucosa with a regenerating epithelium accompanied by ulcer or ulcer scar.
When encountering a fi nding of MPE fi nding at endoscopy, we performed biopsy of both the MPE site and the  non-MPE site. The non-MPE site was defi ned as an adjacent mucosa located within 3 cm of the MPE site (Fig. 2).
The extent of the atrophic border was classifi ed into 3 stages: mild, moderate and severe, as defi ned by Kimura and Takemoto [4].

Histologic assessment
Biopsy specimens were sent to our laboratory, fi xed in 10% buffered formalin. The specimens were processed, embedded in paraffi n, and cut into 4 μm sections. Slides from each specimen were stained using hematoxylin-eosin (HE) (  [10][11][12][13][14]. We used CDX2 because recent studies have shown that CDX2 expression could be observed extensively in IM glands, and did not disappear after eradication of Hp [15,16]. The presence of IM was confi rmed by both HE and immunohistochemical staining. The subtypes of IM were classifi ed as complete or incomplete types. The complete type was defi ned as decreased expression of gastric mucin (MUC5AC or MUC6) and coexpression of intestinal mucin (MUC2) and CD10 (Fig. 3A-F). The incomplete type was defi ned as the expression of gastric mucin (MUC5AC or MUC6) and MUC2. Because no established criteria exist to categorize a case as having both the complete and incomplete types in one section, we assigned the IM subtypes as the prevalent type [15,16].

Statistical analysis
We divided the patients into two groups based on whether  We also compared the histologic fi ndings between the MPE sites and the non-MPE sites. To compare the differences between the two biopsy sites, we used the Wilcoxon Matched-Pairs Signed-Ranks Test for the IM median score, and Fisher's exact test for the prevalence of IM, IM subtypes and IM phenotypes. The sensitivity, specifi city, positive and negative predictive values, and positive and negative likelihood ratios of MPE seen on endoscopy for the detection of pathologic IM were calculated. P values < 0.05 were considered signifi cant. All statistical analyses were performed with Stata software, version 10 (StataCorp LP, College Station, TX, USA).

Patient characteristics
During the study period, 157 patients who underwent up-per endoscopy and received Hp eradication were reviewed. We excluded 52 of the 157 patients from analysis for any of the following criteria: use of antacids (31), anti-thrombotic drugs (17), or NSAIDs (3); or history of liver cirrhosis (5), heart failure (1), or early gastric cancer (2). More than one exclusion criterion applied to some patients.
A total of 102 patients were therefore selected for analysis after exclusion. Of these, there were 55 (54%) patients with MPE. No signifi cant differences were noted in the mean age and sex between MPE-positive and -negative patients ( Table 1). There were no signifi cant differences between the two groups after the Hp eradication period. However, gastric atrophy with a higher severity (moderate to severe) appeared signifi cantly more frequently in the MPE-positive group compared with the MPE-negative group (positive 65.6% versus negative 45.0%, P = 0.035).

Endoscopic fi ndings and pathologic features
The percentage of IM at the MPE sites was signifi cantly higher than at the non-MPE sites (MPE 87.3% versus non-MPE 32.7%, P < 0.001) ( Table 2). The median IM score was signifi cantly higher at the MPE sites than at the non-MPE sites (MPE 2 versus non-MPE 0, P < 0.001) ( Table 2).  Table 2).
No signifi cant associations were observed in the expression of MUC2, MUC5AC, MUC6, CD10, and CDX2 between the MPE sites and the non-MPE sites (Table 3). There was no signifi cant difference in the ratios (complete/incomplete) of IM subtypes between the two groups (MPE: 18/30 versus non-MPE: 8/10, P = 0.778).

Discussion
In this study, we focused on a unique erythematous appearance seen on endoscopy after Hp eradication. We called this fi nding "MPE" and we found that the presence of MPE as seen on endoscopy was typically characteristic of pathologic IM.
The diagnosis of IM with conventional endoscopy has been considered diffi cult because IM usually appears in fl at mucosa and exhibits few morphologic changes. Kaminishi et al reported "ash-colored nodular change" as an indicator for IM; the accuracy of these investigators' fi ndings was high, with a specifi city of 98-99%, but the sensitivity was low (6-12%). Kaminishi et al noted that conventional endoscopy is less useful for confi rming the diagnosis of IM [5]. Recent studies have emerged concerning the endoscopic fi nding of IM using magnifying endoscopy. It has been reported that the distinctive fi ndings of gastric pits seen with methylene blue chromoendoscopy and the "villus-like appearance" seen with confocal endoscopy have been useful for diagnosing IM [17][18][19]. Uedo et al reported that the appearance of "a light blue crest" (LBC) is an accurate sign for the presence of IM, as seen with narrow band imaging (NBI)-magnifying endoscopy [20]. However, due to the high equipment costs and the additional skills and time required for closer examination using such special tools as NBI or magnifying endoscopy, screening with this equipment is not practical in daily clinical practice [21]. In addition, there is an increased risk of damage to the DNA in the gastrointestinal mucosa when using chromoendoscopy with methylene blue followed by white light, requiring caution in its use [22]. Therefore, it is more benefi cial to diagnose IM by fi nding MPE without the use of chromoendoscopy and magnifying endoscopy. Our study results suggest that the presence of IM can be diagnosed with standard endoscopy without biopsy.
Why can MPE be observed even after the eradication of Hp? This could be attributable to the histologic changes within the gastric mucosa. The remarkable histological changes following eradication include improvements in the infi ltration of infl ammatory cells, epithelial hyperplasia and edema [23,24]. The endoscopic images of erythematous and edematous mucosa that appear to be improved refl ect these histologic changes. Therefore, we speculate that MPE consists of a remaining area of persistent erythematous IM and a rapidly recovered non-IM area resulting from successful Hp eradication. The eradication of Hp caused the contrast between the MPE area and non-MPE areas to become clearer. However, it is unknown why only the MPE site is observed as an erythematous mucosa. It can probably be inferred that a highly dense area of microvessels surrounds the metaplastic glands; this has not been elucidated in this study. Additionally, the IM score at the MPE site was signifi cantly higher than the score at the non-MPE site. We speculate that the appearance of erythematous mucosa is associated with the presence of many metaplastic glands.
The present study demonstrated that the IM complete subtype was predominantly found in the gastric mucosa. The subtypes of IM have been classifi ed into either the complete or incomplete type; these are the most widely used subtypes [25]. Several studies have shown that the complete type does not exhibit any increased risk for developing carcinoma, whereas the incomplete type is associated with an increased risk of malignant transformation [26,27]. However, the association between the subtypes and the risk of gastric cancer is not widely accepted [28]. At present, it is diffi cult to identify either of the subtypes using standard endoscopy.
It has been reported that Hp eradication therapy is effective in preventing both gastrointestinal ulcer as well as the development of gastric cancer [29]. These uses of Hp eradication therapy will likely emerge in clinical practice in the near future. However, caution should be taken against the risk of the development of gastric cancer after Hp eradication. The characteristics of gastric cancer after eradication have been reported to include pathologically severe IM at the corpus and severe gastric atrophy as detected on endoscopy [30]. Therefore, it is necessary to carefully observe the presence of IM even after Hp eradication.
In conclusion, the presence of MPE on endoscopic examination is characteristic of pathologic IM. It would be benefi cial in clinical practice to be able to diagnose pathologic IM without chromoendoscopy or magnifying endoscopy. MPE has the potential to become a predictive fi nding for the risk of gastric cancer in the era of Hp eradication.