Incomplete Intestinal Metaplasia in Pyloric Gland Mucosa of Helicobacter pylori‐Naïve Stomachs: A Prospective Multicenter Histopathologic Study Authors: Shinsuke Suemitsu, Kotaro Shibagaki, Asuka Araki, Yoichi Miyaoka, Koichiro Furuta, Tsuyoshi Mishiro, Masahito Aimi, Yusuke Takahashi, Satoshi Kotani, Nobuhiko Fukuba, Norihisa Ishimura, Mamiko Nagase, Shigeru Kawabata, Daisuke Niino, Ryoji Kushima, Shunji Ishihara Published: 2026 ABSTRACT ABSTRACT Helicobacter pylori (Hp)–naïve gastric mucosa has been characterized histologically by the absence of Hp organisms, atrophy, and intestinal metaplasia (IM). We conducted a prospective multicenter study to clarify the prevalence and clinicopathologic features of IM in the Hp‐naïve stomach. Hp‐naïve individuals with regional gastric mucosal abnormalities detected by endoscopy were enrolled. Biopsies were obtained from target lesions and adjacent mucosa. IM frequency and Updated Sydney System (USS) scores were assessed by glandular background (pyloric vs. fundic) in both target and adjacent samples. Among 52 target and 56 adjacent samples, pyloric and fundic gland mucosa accounted for 28 and 24 target samples and 21 and 35 adjacent samples, respectively. IM was detected exclusively in pyloric gland mucosa of target lesions (46.4% vs. 0%, p < 0.05) and was absent in adjacent mucosa. Atrophy scores were low overall but significantly higher in pyloric than fundic gland mucosa in both target lesions (median [IQR], 1 [1–2] vs. 0 [0–1], p < 0.05) and adjacent mucosa (1 [1–2] vs. 1 [0–1], p < 0.05). All IM lesions were classified as incomplete type, showing predominant small intestinal‐type differentiation with residual gastric foveolar markers. Histological evaluation of Hp‐naïve gastric mucosa should be region‐specific, with attention to glandular architecture.Keywords: Helicobacter pylori, intestinal metaplasia, pyloric gland mucosa Abbreviations complete intestinal metaplasia Helicobacter pylori intestinal metaplasia incomplete intestinal metaplasia potassium‐competitive acid blockers proton pump inhibitors spasmolytic polypeptide–expressing metaplasia Updated Sydney System Introduction Gastric cancer is predominantly driven by chronic Helicobacter pylori (Hp)–associated gastritis, typically progressing through atrophic gastritis and intestinal metaplasia (IM) [1, 2]. In Japan, however, the prevalence of Hp infection has markedly declined over recent decades [3], leading to a decrease in Hp‐associated (conventional) gastric cancer and an increasing recognition of Hp‐naïve gastric neoplasm. Histologically, Hp‐naïve gastric mucosa has been characterized by the absence of Hp organisms, minimal inflammatory cell infiltration, and non‐atrophic mucosa without IM, largely based on the Updated Sydney System (USS) [4]. Nevertheless, we previously reported that gastric dysplasia can arise in the pyloric gland mucosa of Hp‐naïve stomachs, with focal IM identified in the background mucosa in approximately 40% of cases and frequently associated with endoscopically detectable mucosal abnormalities [5]. These findings suggest that, in the Hp‐naïve stomach, IM is not a diffuse background alteration but a localized lesion‐associated change arising in a distinct glandular background. Accordingly, we conducted a prospective multicenter study to test the hypothesis that, in the Hp‐naïve stomach, IM preferentially occurs in pyloric gland mucosa rather than fundic gland mucosa. We also investigated the histopathologic and immunophenotypic features of IM identified in this setting. Methods This prospective multicenter observational study enrolled Hp‐naïve individuals with endoscopically detectable regional gastric mucosal abnormalities. The primary endpoint was the IM positivity rate in target lesions. Secondary endpoints were USS scores for neutrophil activity, chronic inflammation, and glandular atrophy in target lesions and adjacent mucosa, the histologic and immunophenotypic characterization of IM‐positive lesions, and comparisons of clinicopathological features and USS scores according to the presence or absence of IM. Participants This study enrolled Hp‐naïve individuals in whom regional gastric mucosal abnormalities (target lesions) were detected during screening upper endoscopy between February 2021 and October 2024. Target lesions included subtle erosions, small depressions, and/or localized color changes—endoscopic findings that may be observed in Hp‐naïve IM [5] but also require differentiation from early gastric cancer. Lesions that could be confidently diagnosed as non‐neoplastic by the endoscopists were excluded. The study protocol was approved by the ethics committees of the five participating institutions, and written informed consent was obtained from all participants. This prospective multicenter observational study was registered in the University Hospital Medical Information Network Clinical Trials Registry (UMIN ID: 000045989), conducted in accordance with the Declaration of Helsinki, and reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement [6]. Diagnostic Criteria of Hp‐Naïve Status Patients were classified as Hp‐naïve if they met all of the following: no history of Hp eradication, negative results in at least two diagnostic tests (serum IgG anti‐Hp antibodies, ¹³C‐urea breath test, fecal Hp antigen test), absence of Hp organisms on histology, and no or minimal gastric mucosal atrophy on endoscopy. A serum IgG anti‐Hp antibody titer of < 3 U/mL was considered truly negative to exclude past infection. The ¹³C‐urea breath test was performed using breath samples collected before and 20 min after ingestion of a 100 mg ¹³C‐urea tablet (UBIT, Otsuka, Japan), measured with an infrared spectrophotometer (POCone, Otsuka Electronics, Japan) at a cutoff of 2.5‰. Clinical Characteristics Patient characteristics, including age, sex, body mass index, smoking status, Brinkman index among smokers, alcohol consumption, arteriosclerotic disease (hypertension, hyperlipidemia, and/or diabetes mellitus), and use of proton pump inhibitors (PPIs) or potassium‐competitive acid blockers (PCABs), were recorded based on prior reports of factors linked to gastric mucosal pathology [7, 8, 9, 10, 11]. Smoking and alcohol consumption were considered present if the patient had a current or past habitual use, defined as use on three or more days per week. The Brinkman index was calculated as the number of cigarettes smoked per day multiplied by the number of years of smoking. Endoscopic Procedure Biopsy specimens were obtained from the target lesion and adjacent mucosa (Figure 1). Lesion location was classified into three regions: the upper or middle gastric body, mainly composed of fundic gland mucosa; the prepyloric region, mainly composed of pyloric glands; and the antral region, where fundic and pyloric glands coexist [12]. The size of each lesion was estimated in millimeters based on the opening width of the biopsy forceps, using conventional biopsy forceps (7 mm) or small‐caliber biopsy forceps (5 mm); the type of forceps used was recorded. Gastric regions and biopsy sampling sites. (A) Lesion location was classified into three regions: the upper (U) and middle (M) gastric body, which are primarily composed of fundic glands; the prepyloric (PP) region, which predominantly consists of pyloric glands; and the antral region, where fundic and pyloric glands coexist. (B) Lesion in the PP region. Biopsy specimens were obtained from the target lesion, identified as a reddish depressed area (yellow arrow), and from the adjacent mucosa (green arrow). (C) Lesion in the U/M region. Biopsy specimens were obtained from the target lesion, identified as a reddish flat area (yellow arrow), and from the adjacent mucosa (green arrow). Histological Analysis Biopsy specimens were fixed in 10% neutral buffered formalin for 24 h, embedded in paraffin, sectioned at 3 μm, and stained with hematoxylin and eosin (H&E). Histological evaluation included the presence of neoplasm, classification of background gastric glands (fundic, pyloric, or predominant type), the presence of IM, and gastritis severity. Gastritis severity was assessed according to the USS, with neutrophil activity, chronic inflammation, and glandular atrophy, each graded as normal (0), mild (1), moderate (2), or marked (3) [4]. All slides were centrally reviewed by three board‐certified pathologists in Shimane University Hospital, and diagnoses were determined by consensus. The pathologists were blinded to the sampling site (target lesion or adjacent mucosa) and the anatomical location of the biopsy. Subtype Classification of IM To differentiate complete IM (CIM) and incomplete IM (IIM), the presence of Paneth cells was evaluated, and immunohistochemical analysis was performed to assess the expression patterns of MUC2 (goblet cell mucin), CD10 (glycosylated transmembrane protein), and CDX2 (intestinal differentiation) as intestinal markers, and MUC5AC (gastric foveolar mucin) and MUC6 (pyloric gland mucin) as gastric markers. CIM was defined by exclusive intestinal marker expression, whereas IIM was defined by a gastrointestinal phenotype, expressing both intestinal markers and at least one gastric marker [13, 14]. Because Paneth cells are located at the gland base and may be under‐sampled in biopsy specimens, particularly when small‐caliber biopsy forceps are used, Paneth cell detection was considered supportive, and final subtype assignment was based on the immunophenotypic profile [15]. Immunohistochemical Analysis Immunohistochemical staining was performed on the Leica Autostainer XL (Leica Biosystems, Germany) following the in‐vitro diagnostic protocol. Sections were incubated with antibodies to MUC2 (clone CCP58, dilution 1:50, DAKO, Glostrup, Denmark), CD10 (clone 56C6, dilution 1:100, DAKO, Glostrup, Denmark), CDX2 (clone EP25, ready‐to‐use, Leica BOND, Nussloch, Germany), MUC5AC (clone CLH2, dilution 1:50, DAKO, Glostrup, Denmark), MUC6 (clone CLH5, dilution 1:200–1:300, Novocastra, Newcastle upon Tyne, UK), gastrin (rabbit polyclonal antibody, dilution 1:2, DAKO, Glostrup, Denmark), and pepsinogen‐I (clone 8003, dilution 1:100, BIO‐RAD, Hercules, CA, USA). Antigen retrieval was performed with ER2 for 20 min for most antibodies, ER1 for pepsinogen‐I, and no retrieval for gastrin. Sample Size Calculation The sample size was based on pilot data. In our preliminary analysis, IM was detected in 37.5% (3/8) of biopsies from Hp‐naïve pyloric gland mucosa with endoscopically detectable mucosal abnormalities, but was absent in fundic gland mucosa. Accordingly, the assumed IM prevalence were set at 37.5% and 0%, respectively, and the required sample size was calculated with a two‐sided α of 0.05 and 80% power, yielding 18 cases per group. Assuming 20% of samples would be inadequate for histologic evaluation, the target sample size was set at 23 cases per group, with paired target and adjacent samples. Statistical Analysis IM positivity rates and USS scores were compared between pyloric gland mucosa and fundic gland mucosa in both target lesions and adjacent mucosa. Clinicopathological features and USS scores were also compared according to the presence or absence of IM. Continuous variables were expressed as means ± standard deviations. Ordinal variables, including USS scores, were expressed as medians with interquartile ranges (IQR), whereas nominal categorical variables were presented as numbers and percentages. Univariate analyses were performed using Student's or Welch's t‐test, or the Mann–Whitney U test, for continuous variables, and the chi‐square test or Fisher's exact test for nominal categorical variables. Statistical significance was set at p < 0.05. Results A total of 61 participants with 62 lesions were enrolled. Lesion distribution was as follows: 20 in the upper or middle gastric body, 8 in the antrum, and 34 in the prepyloric region. Based on biopsy findings, 2 patients (2 lesions) were excluded because of a final diagnosis of intestinal‐type dysplasia or signet‐ring cell carcinoma (Supporting Figure 1). In addition, cases in which biopsy specimens contained only foveolar epithelium without glandular components were excluded, resulting in the exclusion of target samples from 8 patients (8 lesions) and adjacent samples from 4 patients (4 lesions). Consequently, the final analysis included target lesions from 51 patients with 52 lesions and adjacent mucosa from 55 patients with 56 lesions, as summarized in Supporting Figure 2. Clinicopathologic Characteristics of Pyloric and Fundic Gland Mucosa in Target Lesions and Adjacent Mucosa In target lesions, pyloric gland mucosa was identified in 28 lesions from 27 patients and fundic gland mucosa in 24 lesions from 24 patients; in adjacent mucosa, pyloric gland mucosa was identified in 21 lesions from 20 patients and fundic gland mucosa in 35 lesions from 35 patients. Baseline characteristics are shown in Table 1. Although the fundic gland mucosa group was older, other variables were comparable between groups. Table 1Clinical characteristics of target lesions in pyloric and fundic gland mucosa.Pyloric gland mucosa (27 patients, 28 lesions)Fundic gland mucosa (24 patients, 24 lesions) p valueAge, mean ± SD, years49.4 ± 7.457.6 ± 11.1< 0.05Sex (male/female), n 17/1011/130.22BMI, mean ± SD, kg/m2 23.0 ± 2.823.2 ± 4.00.78Smoking habit, % (n/N)37.0 (10/27)33.3 (8/24)0.78Brinkman Index, mean ± SD112.0 ± 196.7162.6 ± 281.50.93Alcohol consumption, % (n/N)44.4 (12/27)37.5 (9/24)0.61Arteriosclerotic disease, % (n/N)25.6 (7/27)16.7 (4/24)0.42PPI or PCAB use, % (n/N)7.4 (2/27)16.7 (4/24)0.28Type of biopsy forceps (conventional/small), n 15/1314/100.73Abbreviations: BMI, body mass index; PCAB, potassium‐competitive acid blocker; PPI, proton pump inhibitor; SD, standard deviation. Clinical characteristics of target lesions in pyloric and fundic gland mucosa. Abbreviations: BMI, body mass index; PCAB, potassium‐competitive acid blocker; PPI, proton pump inhibitor; SD, standard deviation. Histologically, in target lesions (Figure 2a), neutrophil activity and chronic inflammation scores (median [interquartile range]) did not differ between pyloric and fundic gland mucosa (0 [0–1] vs. 0 [0–1], p = 0.8; and 1 [1–2] vs. 1 [1–1], p = 0.16, respectively). In contrast, although low overall, the atrophy score was slightly but significantly higher in pyloric gland mucosa (1 [1–2] vs. 0 [0–1], p < 0.05). IM was detected only in pyloric gland mucosa (46.4% vs. 0%, p < 0.05). Consistently, the intestinal metaplasia score differed significantly between pyloric and fundic gland mucosa (0 [0–1] vs. 0 [0–0], p < 0.05). In adjacent mucosa (Figure 2b), neutrophil activity and chronic inflammation scores were similarly low in pyloric and fundic gland mucosa (0 [0–0] vs. 0 [0–0], p = 0.59; and 1 [1–1] vs. 1 [1–1], p = 0.83, respectively). Atrophy scores were low in both mucosal types but showed a slight shift toward higher values in pyloric gland mucosa (1 [1–2] vs. 1 [0–1], p < 0.05), whereas IM was not observed in either mucosa. USS scores of neutrophil activity, chronic inflammation, atrophy, and intestinal metaplasia in target lesions (A) and adjacent mucosa (B) in Helicobacter pylori–naïve stomachs, stratified by background mucosal type (pyloric and fundic gland mucosa). Sample sizes were 28 and 24 for pyloric and fundic gland mucosa, respectively, in target lesions (A), and 21 and 35, respectively, in adjacent mucosa (B). Dot plots show individual USS scores, and horizontal bars indicate medians. p values were calculated using the Mann–Whitney U test. N/A, not applicable. Clinical Characteristics of Target Lesions in Pyloric Gland Mucosa With and Without Intestinal Metaplasia Table 2 summarizes the clinical characteristics of target lesions in the pyloric gland mucosa, including 13 patients (13 lesions) with IM and 14 patients (15 lesions) without IM. The majority of the lesions were located in the prepyloric region regardless of the presence or absence of IM. Most IM‐positive lesions were reddish and depressed. Depressed morphology was numerically more frequent in IM‐positive than IM‐negative lesions (9/13 [69.2%] vs. 4/15 [26.7%]); however, the overall distributions of endoscopic color and morphology did not differ significantly between the two groups. There were no significant differences between the two groups in age, sex, BMI, or the use of PPI or PCAB. Although not statistically significant, the IM‐positive group showed a numerically higher prevalence of smoking (46.2% vs. 21.4%, p = 0.17) and a greater mean Brinkman index (178.1 vs. 50.7, p = 0.07). Arteriosclerotic disease was significantly higher in the IM‐positive group than in the IM‐negative group (46.2% vs. 7.1%, p < 0.05). Table 2Clinical characteristics of target lesions in pyloric gland mucosa with and without intestinal metaplasia.Intestinal metaplasia p valuePresence (13 patients, 13 lesions)Absence (14 patients, 15 lesions)Age, mean ± SD, years52.3 ± 7.946.2 ± 6.30.18Sex (male/female), n 9/48/60.4BMI, mean ± SD, kg/m2 23.8 ± 3.522.2 ± 1.70.15Smoking habit, % (n/N)46.2 (6/13)21.4 (3/14)0.17Brinkman Index, mean ± SD178.1 ± 241.950.7 ± 122.20.07Alcohol consumption, % (n/N)30.8 (4/13)57.1 (8/14)0.17Arteriosclerotic disease, % (n/N)46.2 (6/13)7.1 (1/14)< 0.05PPI or PCAB use, % (n/N)7.7 (1/13)7.1 (1/14)0.74Type of biopsy forceps (conventional/small), n 7/68/70.97Macroscopic diameter, mean ± SD, mm5.3 ± 3.87.4 ± 4.20.33Location (PP/Antrum/UM)11/2/014/1/00.44Color (reddish/yellowish/isocolor)10/1/213/1/10.78Morphology (erosion/depression/elevation)2/9/26/4/50.11Abbreviations: BMI, body mass index; PCAB, potassium‐competitive acid blocker; PPI, proton pump inhibitor; PP, prepyloric region; SD, standard deviation; UM, upper or middle gastric body. Clinical characteristics of target lesions in pyloric gland mucosa with and without intestinal metaplasia. Abbreviations: BMI, body mass index; PCAB, potassium‐competitive acid blocker; PPI, proton pump inhibitor; PP, prepyloric region; SD, standard deviation; UM, upper or middle gastric body. Histological Features of Target Lesions in Pyloric Gland Mucosa With and Without Intestinal Metaplasia A representative case with IM is shown in Figure 3. The histological scores of neutrophil activity, chronic inflammation, and atrophy of target lesions in pyloric gland mucosa with or without IM are summarized in Figure 4. Neutrophil activity scores (median [interquartile range]) were 1 [0–1] vs. 0 [0–0] (p = 0.09) in the IM‐positive and IM‐negative groups, respectively, suggesting a trend toward slightly higher activity in the IM‐positive group. Chronic inflammation scores also showed a shift toward higher values in the IM‐positive group, with values of 1 [1–2] vs. 1 [1–1] (p = 0.12). In contrast, atrophy scores were comparable between the two groups, with values of 1 [1–2] vs. 1 [1–2] (p = 0.49). In the IM‐negative group, the lesions were characterized mainly by foveolar hyperplasia and lamina propria edema, with scant inflammatory cell infiltration. Endoscopic and histological findings of a Helicobacter pylori‐naïve case with intestinal metaplasia (IM). (A) A slightly depressed, reddish lesion in the prepyloric region. Biopsies were obtained from the target lesion (blue arrow) and the adjacent mucosa. (B) Low‐power view of the biopsy specimen from the target lesion (H&E, ×40). (C) Higher‐power view of the area indicated in (B), showing IM in the background pyloric gland mucosa (H&E, ×100). (D) Paneth cells at the base of IM glands (yellow arrow) (H&E, ×200). USS scores were 0 for neutrophil activity, 1 for chronic inflammation, and 2 for atrophy. USS scores of neutrophil activity, chronic inflammation, and atrophy in Helicobacter pylori–naïve pyloric gland mucosa of target lesions with IM (IM‐present, n = 13) and without IM (IM‐absent, n = 15). Dot plots show individual USS scores, and horizontal bars indicate medians. p values were calculated using the Mann–Whitney U test. Histological Features of Intestinal Metaplasia in Helicobacter pylori‐Naïve Pyloric Gland Mucosa Representative histological and immunohistochemical findings of IM in 13 cases from the Hp‐naïve pyloric gland mucosa are shown in Figure 5. H&E staining demonstrated the presence of Paneth cells in 76.9% (10/13) of cases. Immunohistochemically, all lesions exhibited strong nuclear expression of CDX2 and diffuse cytoplasmic expression of MUC2 confined to goblet cells. CD10 was positive in 84.6% (11/13). In contrast, gastric‐type markers showed sporadic expression of MUC5AC in all cases, whereas MUC6 showed weak and focal positivity at the gland base of IIM in a subset of cases. Staining for MUC5AC and CD10 was heterogeneous (mosaic) within the metaplastic epithelium and tended to show a reciprocal distribution at the glandular level, with focal residual MUC5AC positivity observed even in some CD10‐positive ducts. Accordingly, all lesions were classified as IIM and predominantly exhibited a small intestinal‐type differentiation with residual gastric foveolar marker. Histological and immunohistochemical features of intestinal metaplasia (IM) in pyloric gland mucosa of a Helicobacter pylori‐naïve stomach. (A) IM with diffusely distributed goblet cells (H&E, ×200). (B) Paneth cells within the IM epithelium (yellow arrow) (H&E, ×400). (C) Partial expression of the gastric marker MUC5AC (×200). Intestinal markers showed strong nuclear CDX2 expression (D, ×200), diffuse cytoplasmic MUC2 expression (E, ×200), and apical membranous CD10 expression (F, ×200), indicating a predominantly small intestinal phenotype with residual gastric foveolar differentiation. Panels C–F are from serial sections. MUC5AC and CD10 staining was heterogeneous (mosaic) and tended to show a reciprocal gland‐level distribution, with focal residual MUC5AC positivity observed even in CD10‐positive ducts. Background Pyloric Gland Mucosa in Intestinal Metaplasia‐Positive Lesions The background mucosa of all IM‐positive target lesions consisted of pyloric gland mucosa and consistently showed gastrin‐producing cells with absent pepsinogen I expression (Supporting Figure 3), suggesting that the background mucosa represented native pyloric gland mucosa rather than pyloric or pseudopyloric gland metaplasia. Discussion The present study is the first to investigate the prevalence and histopathologic features of IM in the Hp‐naïve stomach, showing that IM frequently arises in the pyloric gland mucosa. IM was detected in nearly half of the target lesions in the pyloric gland mucosa, but in neither the fundic gland mucosa nor the adjacent mucosa. In contrast, atrophy scores were slightly but consistently higher in the pyloric gland mucosa than in the fundic gland mucosa in both target lesions and adjacent mucosa. These findings indicate that, in the Hp‐naïve stomach, the pyloric gland mucosa may show mild atrophy, whereas IM is confined to endoscopically detectable mucosal abnormalities within it. Importantly, the Hp‐naïve gastric IM in our cohort was uniformly classified as IIM, and immunohistochemistry revealed predominantly small intestinal–type differentiation while retaining gastric foveolar features. MUC5AC and CD10 showed a mosaic staining pattern within the metaplastic epithelium and were largely reciprocally distributed, with focal residual MUC5AC positivity observed even in CD10‐positive ducts. A previous study of Hp‐exposed gastric IM showed that IIM with a gastrointestinal phenotype was characterized by persistence of gastric foveolar mucin within goblet and absorptive cells, accompanied by weak villin expression [16]. In experimental injury models, IIM predominates early, followed by a relative increase in CIM over time [17]. Clinical data after Hp eradication have also suggested plasticity of IM subtypes, including a decrease in IIM over time, although direct histological evidence that IIM can revert to an antecedent gastric mucosal state has not been established [15]. These findings support the concept that gastric IIM represents an intermediate stage of intestinal differentiation rather than discrete entities [18]. In this context, the absence of CIM in our Hp‐naïve cohort may suggest a pathogenesis distinct from the conventional Hp‐associated Correa cascade [19], with limited progression toward complete intestinalization. The evidence base for IM subtype‐guided risk stratification, including meta‐analytic data suggesting a higher risk of gastric neoplasia in IIM than in CIM, has been established in Hp‐exposed cohorts [20, 21], but its applicability to Hp‐naïve IM remains uncertain. Among Hp‐naïve gastric neoplasms, intestinal‐type dysplasia is a dysplastic intestinal epithelial lesion arising in pyloric gland mucosa and immunohistochemically shows an intestinal or gastrointestinal phenotype. In some cases, IM is sporadically observed around these lesions (Supporting Figure 4) [5]. Further studies are needed to clarify whether Hp‐naïve IIM may represent a potential tumorigenic background for this lesion. In the Hp‐infected stomach, IM develops through the stepwise inflammatory process [19, 22]. In the Hp‐naïve pyloric gland mucosa examined in this study, neutrophil activity and chronic inflammation scores tended to be higher in IM‐positive lesions, whereas atrophy scores were comparable regardless of IM status, suggesting that Hp‐independent mucosal injury and inflammation may contribute to IM development in this setting. IM‐positive cases tended to show a higher prevalence of smoking and arteriosclerotic disease, both of which have been associated with chronic epithelial injury, impaired mucosal perfusion, and hypoxia‐related metaplastic change [23, 24, 25, 26, 27, 28]. In addition, bile reflux has also been implicated in IM development via bile acid–mediated signaling and epigenetic regulation [29, 30, 31], and prior data suggest that IIM, compared with CIM, may be more closely linked to bile reflux–associated settings [32, 33]. Smoking has been reported to be associated with increased bile reflux [34], and other recognized contributors to primary biliary reflux include cholecystectomy, gallbladder dysfunction, biliary sphincterotomy, and gastric or duodenal dysmotility [35]. Histologically, bile reflux–associated injury may show features of chemical/reactive gastropathy, including foveolar hyperplasia, lamina propria edema, smooth muscle fibers in the lamina propria, and paucity of inflammatory cells [36, 37]. However, these findings are not specific to bile reflux and may also occur in other forms of chemical injury, including that related to nonsteroidal anti‐inflammatory drugs [35]. Because bile reflux was not directly assessed in the present study, the possible contribution of bile‐related mucosal injury to Hp‐naïve IM remains speculative. An important consideration in this study was whether IIM arose from native pyloric glands or reflected secondary pyloric/pseudopyloric metaplasia after oxyntic gland injury. Spasmolytic polypeptide–expressing metaplasia (SPEM) is a pyloric‐type metaplastic mucous cell lineage associated with oxyntic atrophy and characterized by MUC6/TFF2 expression [38, 39]. SPEM has been proposed to reflect reprogramming of mature chief cells toward pyloric‐type mucous cells after oxyntic injury. However, in our series, IIM‐positive pyloric gland mucosa showed overall low‐grade inflammatory activity, preserved gastrin‐positive cells, and no pepsinogen I expression, supporting IIM development within native pyloric glands in the Hp‐naïve stomach. Barrett's esophagus provides a relevant model of stepwise metaplastic change in an Hp‐naïve setting. Serial biopsy studies in reflux models suggest that cardiac‐type columnar mucosa precedes specialized columnar epithelium, corresponding to IM, supporting a precursor role of cardiac‐type mucosa [40, 41, 42], and IIM is the predominant phenotype in Barrett's esophagus [43, 44]. Given the morphological and functional similarity between pyloric and cardiac glands, these observations support the notion that pyloric‐ or cardiac‐type glandular epithelium provides a permissive lineage context for IM, particularly IIM, development. Conversely, in Hp‐associated gastritis, IM in the oxyntic compartment is frequently observed within areas of pyloric/pseudopyloric metaplasia [45], suggesting that establishment of a pyloric‐type mucous cell lineage may be a prerequisite for subsequent IM development. This study has several limitations. First, biopsies were obtained from endoscopically detected regional mucosal abnormalities, introducing inherent selection bias; therefore, the observed IM prevalence may not represent the overall Hp‐naïve stomach. Second, without systematic mapping biopsies, the stomach‐wide extent of IM could not be fully defined; however, IM in the endoscopically non‐atrophic oxyntic compartment appears highly unlikely in our cohort. Finally, this cross‐sectional study did not assess longitudinal outcomes, including the risk of subsequent neoplasm. In conclusion, the Hp‐naïve stomach may exhibit mild atrophy in the pyloric gland mucosa, whereas IM is confined to endoscopically detectable mucosal abnormalities. Histologic evaluation of Hp‐naïve gastric mucosa should take the underlying glandular architecture into account. Author Contributions Conception and design: Shinsuke Suemitsu and Kotaro Shibagaki. Conception and design: Shinsuke Suemitsu and Kotaro Shibagaki. Analysis and interpretation of data: Shinsuke Suemitsu, Kotaro Shibagaki, Asuka Araki, Mamiko Nagase, Satoshi Kotani, Daisuke Niino. Analysis and interpretation of data: Shinsuke Suemitsu, Kotaro Shibagaki, Asuka Araki, Mamiko Nagase, Satoshi Kotani, Daisuke Niino. Patient enrollment: Shinsuke Suemitsu, Kotaro Shibagaki, Yoichi Miyaoka, Koichiro Furuta, Tsuyoshi Mishiro, Masahito Aimi, Yusuke Takahashi, Nobuhiko Fukuba. Patient enrollment: Shinsuke Suemitsu, Kotaro Shibagaki, Yoichi Miyaoka, Koichiro Furuta, Tsuyoshi Mishiro, Masahito Aimi, Yusuke Takahashi, Nobuhiko Fukuba. Critical revision of manuscript for important intellectual content: Norihisa Ishimura, Ryoji Kushima. Critical revision of manuscript for important intellectual content: Norihisa Ishimura, Ryoji Kushima. Final approval of the study: Shunji Ishihara. Final approval of the study: Shunji Ishihara. Ethics Statement The protocol of this study was approved by the medical ethics committee of Shimane University Hospital (Study number: 6520, Date: March 10, 2021). Conflicts of Interest Ryoji Kushima is an Associate Editor of Pathology International and a co‐author of this article. The remaining authors declare no conflicts of interest. Supporting information Figure S1: Representative excluded cases in which biopsy established a definitive diagnosis of Helicobacter pylori‐naïve gastric neoplasia and subsequent endoscopic resection achieved curative resection. Intestinal‐type gastric dysplasia: (A) White‐light endoscopy showing a small depressed lesion with marginal elevation, mimicking verrucous gastritis (yellow arrow). (B) Histology showing dysplastic tubular glands. Signet‐ring cell carcinoma: (C) White‐light endoscopy showing a shallow depressed lesion (red arrows). (D) Histology showing diffuse infiltration of signet‐ring cells in the lamina propria. Figure S2: Flow diagram of case selection and specimen eligibility. Figure S3: Background mucosa of a target lesion located in the Helicobacter pylori‐naïve prepyloric region in a case with intestinal metaplasia (same case as Figure 3). (A) Background mucosa composed of pyloric gland mucosa (H&E, × 100). Immunohistochemical staining demonstrated diffuse expression of MUC6 in the pyloric glands (B, × 100), presence of gastrin‐positive cells in the gland neck region (C, × 100), and absence of pepsinogen I expression (D, × 100). Figure S4: Endoscopic and histological features of intestinal‐type dysplasia adjacent to sporadic intestinal metaplasia (IM) in the prepyloric area of a Helicobacter pylori–naïve stomach. (A) White‐light endoscopy revealed irregularly shaped multiple depressed lesions with slightly elevated surrounding mucosa. Histological mapping of the resected specimen (B) and the corresponding endoscopic image (C); yellow lines indicate intestinal‐type dysplasia, dashed line indicates the area where dysplasia was covered by non‐neoplastic epithelium, and white lines indicate IM. (D) Low‐power view of the segment indicated by the red dashed line in (B). (E) Higher‐power view of the red‐boxed area in (D), showing irregular tubular glands with goblet cell‐like mucin (yellow arrows) adjacent to IM arising from the proliferative zone of pyloric gland mucosa (H&E, × 100). (F) Higher‐power view of the blue‐boxed area in (E) (H&E, × 200). Acknowledgments We sincerely thank Dr. Kazuyoshi Yagi (Niigata Association of Occupational Health) for his valuable clinical guidance and constructive advice regarding this study. Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. References Associated Data Supplementary Materials Figure S1: Representative excluded cases in which biopsy established a definitive diagnosis of Helicobacter pylori‐naïve gastric neoplasia and subsequent endoscopic resection achieved curative resection. Intestinal‐type gastric dysplasia: (A) White‐light endoscopy showing a small depressed lesion with marginal elevation, mimicking verrucous gastritis (yellow arrow). (B) Histology showing dysplastic tubular glands. Signet‐ring cell carcinoma: (C) White‐light endoscopy showing a shallow depressed lesion (red arrows). (D) Histology showing diffuse infiltration of signet‐ring cells in the lamina propria. Figure S2: Flow diagram of case selection and specimen eligibility. Figure S3: Background mucosa of a target lesion located in the Helicobacter pylori‐naïve prepyloric region in a case with intestinal metaplasia (same case as Figure 3). (A) Background mucosa composed of pyloric gland mucosa (H&E, × 100). Immunohistochemical staining demonstrated diffuse expression of MUC6 in the pyloric glands (B, × 100), presence of gastrin‐positive cells in the gland neck region (C, × 100), and absence of pepsinogen I expression (D, × 100). Figure S4: Endoscopic and histological features of intestinal‐type dysplasia adjacent to sporadic intestinal metaplasia (IM) in the prepyloric area of a Helicobacter pylori–naïve stomach. (A) White‐light endoscopy revealed irregularly shaped multiple depressed lesions with slightly elevated surrounding mucosa. Histological mapping of the resected specimen (B) and the corresponding endoscopic image (C); yellow lines indicate intestinal‐type dysplasia, dashed line indicates the area where dysplasia was covered by non‐neoplastic epithelium, and white lines indicate IM. (D) Low‐power view of the segment indicated by the red dashed line in (B). (E) Higher‐power view of the red‐boxed area in (D), showing irregular tubular glands with goblet cell‐like mucin (yellow arrows) adjacent to IM arising from the proliferative zone of pyloric gland mucosa (H&E, × 100). (F) Higher‐power view of the blue‐boxed area in (E) (H&E, × 200). Data Availability Statement The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.