Subtype-specific risk factors for gastric adenomas and independent predictors of recurrence after endoscopic resection

Abstract

BACKGROUND

Gastric adenoma is widely acknowledged as a premalignant lesion that can progress to gastric adenocarcinoma. The distinctions among subtypes of gastric adenomas have been infrequently explored in existing literature.

AIM

To assess the subtype-specific risk factors and outcomes of endoscopic resection (ER) for gastric adenomas.

METHODS

This is a retrospective cohort study. Among 162 patients who underwent ER for gastric lesions larger than 10 mm between 2017 and 2022, 79 patients with gastric adenomas were included. Hyperplastic polyps (n = 37) and subepithelial lesions (n = 46) were excluded. Logistic regression and survival analyses were conducted.

RESULTS

The 79 patients (mean age 68.1 years; 65% male) had adenoma subtypes: 62% intestinal, 29% foveolar, 8% pyloric, and 1% oxyntic. The mean follow-up was 26 months. Intestinal adenoma was strongly linked to a family history of gastric adenocarcinoma and atrophic gastritis (P < 0.001); foveolar adenoma was significantly associated with intestinal metaplasia (P < 0.001). Pyloric adenomas had the largest polyp size (P < 0.001). Recurrence rates were 8%, 17%, and 17% for the respective subtypes (P = 0.07), with no significant difference in the meantime to recurrence (P = 0.8). Independent predictors of recurrence after ER included a family history of gastric adenocarcinoma, active Helicobacter pylori infection, polyp size ≥ 30 mm, presence of > 3 polyps and Paris 0-IIc morphology (P < 0.001).

CONCLUSION

This study identifies distinct risk factor profiles for different subtypes of gastric adenomas and independent recurrence predictors post-ER, underscoring the importance of subtype-specific tailored risk assessment and surveillance strategies.

Key Words: Gastric adenoma; Intestinal adenoma; Foveolar adenoma; Pyloric adenoma; Gastric adenoma recurrence; Helicobacter pylori; Endoscopic submucosal dissection

Core Tip: Gastric adenomas exhibit subtype-specific characteristics; intestinal adenomas linked to family history of gastric adenocarcinoma and atrophic gastritis, foveolar adenomas associated with nonsteroidal anti-inflammatory drug use and intestinal metaplasia, and pyloric adenomas presenting as large multiple polyps. Understanding the unique clinicopathological features of different subtypes of gastric adenomas helps to implement risk-stratified surveillance strategies and address modifiable factors like Helicobacter pylori infection to reduce recurrence rates.

INTRODUCTION

Gastric adenoma is widely acknowledged as a premalignant lesion that can progress to gastric adenocarcinoma[1]. Much of the existing literature groups all adenoma subtypes together, categorizing them based on features such as tubular, villous, or mixed structures, as well as flat or depressed morphologies[1-3]. However, gastric adenomas can be classified into four distinct types. Intestinal adenomas, which account for over half of all gastric adenomas, are characterized by foci of intestinal epithelium and are often associated with atrophic gastritis and intestinal metaplasia[4]. The foveolar type, the second most common subtype, exhibits dysplasia but is generally considered less aggressive than intestinal adenomas and is rarely linked to high-grade dysplasia (HGD) or adenocarcinoma[5]. Foveolar adenomas typically develop in normal mucosal tissue and are frequently associated with familial adenomatous polyposis (FAP)[4]. Pyloric-type adenomas, similar to intestinal-type adenomas, are associated with an aggressive clinical course and arise in the setting of atrophic gastritis[6]. Oxyntic gland adenomas, previously regarded as rare precursors to gastric adenocarcinoma with specific chief cell differentiation, have not been reported to recur or metastasize to date[6].

The diagnosis of gastric adenoma is typically made through endoscopic biopsy sampling. However, due to discrepancies between biopsy results and findings from resected specimens, complete endoscopic resection (ER) of all gastric adenomas is recommended[7-10]. The incidence of synchronous or metachronous recurrence after ER ranges from 3% to 20.9%[11], necessitating mandatory endoscopic surveillance at three months post-resection and annually thereafter[12,13]. To our knowledge, few Western studies have comprehensively investigated gastric adenomas[12,14]. This study aims to evaluate the prevalence, potential risk factors, and outcomes of endoscopic management for the different subtypes of gastric adenoma.

MATERIALS AND METHODS

Study design and patient variables

This retrospective cohort study was conducted at a French tertiary referral center for digestive endoscopy. Data were obtained from a prospectively collected database of consecutive patients who underwent ER for gastric lesions between 2017 and 2022. The data collected included patient demographics, body mass index (BMI), history of comorbidities, family history of cancer, personal history of gastric neoplasia, history of drug intake [statin, proton pump inhibitor (PPI), nonsteroidal anti-inflammatory drug (NSAID) and aspirin], and endoscopic (number, size and location) and histopathological features [histological subtype, chronic gastritis, Helicobacter pylori (H. pylori) infection]. Data regarding the type of endoscopic treatment, either endoscopic submucosal dissection (ESD) or endoscopic mucosal resection (EMR), and post ER recurrence (location, number, size, histological type and time to recurrence after resection) were also collected.

Ethics statement

Data were obtained from electronic patient records following approval by the local hospital ethics committee (reference number: AAA-2023-09011). The study was conducted in accordance with the Declaration of Helsinki and adhered to the principles of good clinical practice.

ER techniques

At our endoscopy unit, ER techniques were tailored according to lesion characteristics and subtype. Intestinal adenomas (often antral, with atrophic gastritis) required careful complete resection margins due to adenocarcinoma risk. Foveolar adenomas (frequently multiple, Paris 0-IIa+IIc) necessitated meticulous mapping and staged R0 resection. Larger, Paris 0-Ippyloric adenomas often require ESD.

Magnifying endoscopy with narrow-band imaging (NBI) was utilized to assess micro-surface patterns, particularly for lesions with indeterminate borders. Foveolar adenomas typically demonstrated regular micro-surface patterns with fine granular or homogeneous appearances, while intestinal adenomas often showed irregular micro-surface patterns consistent within intestinal metaplasia with disrupted microvascular architecture. Pyloric adenomas exhibited characteristic tubular or papillary micro-surface patterns with regular microvascular patterns. Oxyntic adenomas, represented by a single case, showed homogeneous micro surface patterns with regular microvascular architecture[15,16].

Delineating resection margins for foveolar adenomas is challenging due to their subtle endoscopic appearance and similarity to hyperplastic polyps. To optimize R0 resection, we employed careful pre-resection assessment using magnifying NBI to identify demarcation lines, ensuring a 2-3 mm safety margin. En bloc resection was performed when feasible (especially for lesions > 20 mm), while larger lesions underwent piecemeal resection with precise margin mapping. Post-resection assessment with white light and NBI confirmed complete removal, with further resection if suspicious margins persisted[17].

Follow-up after ER

The follow-up period was calculated from the date of ER to the last follow-up visit, the date of death, or the study's conclusion (January 31, 2024). All patients were scheduled for follow-up endoscopic evaluations at 3-or 6-months post-ER, followed by additional examinations every 6 months and annually thereafter.

Residual and recurrent adenoma

Residual adenoma was identified as the presence of a new adenoma detected within the first 3-6 months following ER. In contrast, recurrent adenoma was defined as the detection of an adenoma occurring more than 6 months after the last ER. Recurrence could manifest either at the same location as the previously resected adenoma (local recurrence) or at a different site distant from the primary adenoma (metachronous recurrence).

Statistical analysis

All the statistical analyses were performed using the Statistical Package for the Social Sciences version 26 (SPSS, Inc., Chicago, IL, United States) and R software version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). Categorical variables are presented as percentages and were compared via the χ² test with Fisher’s correction. Continuous variables are expressed as the means with SD and were compared via ANOVA. Recurrence and residual rates were assessed via the Kaplan-Meier method and compared via the log-rank test. Statistical significance was accepted with a P value ≤ 0.05 and 95% confidence interval (CI). A multivariable logistic regression model was applied to identify potential predictors for the histological subtypes. The results of the multivariable logistic model are expressed as odds ratios (OR) with P value. Multivariate Cox regression in a backward manner was designed to determine the effect of all the statistically significant variables with P values ≤ 0.05 for recurrence outcome. The results of the multivariable Cox model are expressed as hazard ratios (HR) with P values. Model β estimates were used to calculate HRs and risk scores for predicting recurrent/residual adenoma.

RESULTS

Patient characteristics

Of the 162 patients who underwent endoscopic ER for gastric lesions between 2017 and 2022, 79 were included in this study following diagnosis of gastric adenoma. Patients with hyperplastic polyps (n = 37) or subepithelial lesions (n = 46) were excluded. The 79 patients were classified into four groups based on histological subtypes of gastric adenoma: Intestinal (n = 49), foveolar (n = 23), pyloric (n = 6), and oxyntic (n = 1) (Figure 1).

Figure 1  The study flow diagram.

The clinical characteristics of these 79 patients at the time of gastric adenoma diagnosis is summarized in Table 1. Patients with oxyntic adenoma were excluded from statistical analysis to avoid bias. The majority of patients were male (64.6%) with a mean BMI of 24.9 ± 3.9 kg/m². Cardiovascular diseases were the most common comorbidity (58.2%), followed by malignancies (27.8%) and diabetes mellitus (21.5%). Histories of PPI and NSAID/aspirin intake were recorded in 94.9% and 68.4% of patients, respectively (Table 1).

Table 1 Patients’ baseline characteristics (n = 79), mean ± SD.

	n (%)
Gender
Male	51 (64.6)
Female	28 (35.4)
Mean age, years	68.1 ± 14.6
Mean BMI, kg/m2	24.9 ± 3.9
Comorbidities
Cardiovascular diseases	46 (58.2)
Malignancy	22 (27.8)
Diabetes	17 (21.5)
Chronic kidney disease	4 (5.1)
Morbid obesity	3 (3.8)
Cirrhosis	2 (2.5)
History of PPI intake	75 (94.9)
History of NSAID/Aspirin intake	54 (68.4)
History of alcohol abuse	39 (49.4)
History of smoking	34 (43.0)
History of statin intake	26 (32.9)
Hereditary predisposition
FAP	4 (5.1)
Peutz Jeghers/juvenile polyposis	2 (2.5)
Others	2 (2.5)
Personal history of gastric neoplasia
Adenocarcinoma	14 (17.7)
Adenoma	7 (8.9)
Family history
Gastric adenocarcinoma	8 (10.1)
Colorectal cancer	5 (6.3)
Other cancers	3 (3.8)

BMI: Body mass index; NSAID: Nonsteroidal anti-inflammatory drug; PPI: Proton pump inhibitor; FAP: Familial adenomatous polyposis.

Eight patients had hereditary predispositions to gastric neoplasia: Four (5.1%) with FAP, two (2.5%) with juvenile polyposis, and two (2.5%) with other hereditary predispositions. Personal histories included gastric adenocarcinoma in 14 patients (17.7%) and gastric adenoma in seven patients (8.9%) (Table 1).

Endoscopic and histopathological characteristics

The majority of the patients (86.1%) had a single lesion. The mean size of the resected specimens was 31.9 ± 19.2 mm. The most common location for gastric adenoma was the antrum (36.7%), followed by the body (greater curvature: 22.8% and lesser curvature: 10.1%). Approximately 27.8% of the adenomas were Paris classification 0-IIa, and 25.4% of them were Paris classification 0-IIa+IIc. Half of the lesions (51.9%) were reddish in color. Most of the lesions (82.3%) were resected en bloc via ESD, whereas 17.7% were resected via EMR. Histologically complete ER was achieved in 73 (92.4%) patients (Table 2).

Table 2 Baseline endoscopic and histopathological characteristics (n = 79), mean ± SD.

	n (%)
Baseline endoscopic characteristics:
Number of lesions
1/2/≥ 3	68 (86.1)/5 (6.3)/6 (7.6)
Mean size, mm	31.9 ± 19.2
Location
Cardia/fundus	21 (26.6)
Greater curvature	18 (22.8)
Lesser curvature	8 (10.1)
Antrum	29 (36.7)
Pylorus	3 (3.8)
Gross morphology (Paris classification)
0-Ip	10 (12.7)
0-Is	17 (21.5)
0-IIa	22 (27.8)
0-IIb	2 (2.5)
0-IIc	6 (7.6)
0-IIa+IIc	20 (25.4)
0-Is+IIc	2 (2.5)
Color
Reddish	41 (51.9)
Same as background mucosa	20 (25.3)
Ulcerated/erosive	16 (20.3)
Whitish	2 (2.5)
Treatment
EMR	14 (17.7)
ESD	65 (82.3)
Histopathological characteristics:
Histological subtype
Intestinal	49 (62.0)
Foveolar	23 (29.1)
Pyloric	6 (7.6)
Oxyntic	1 (1.3)
Helicobacter pylori infection
Negative	56 (70.9)
Past infection	18 (22.8)
Current infection	4 (5.1)
Mean size of resected specimen, mm	37.7 ± 34.3
Underlying stomach
Normal	11 (13.9)
Atrophic gastritis	39 (49.4)
Intestinal metaplasia	29 (36.7)
Worst histology
LGD	22 (27.8)
HGD	18 (22.8)
Intramucosal adenocarcinoma (T1a)	21 (26.6)
Submucosal carcinoma (T1b)	18 (22.8)
Resection completeness
Histologically complete (R0)	73 (92.4)
Piecemeal	5 (6.3)
Failed endoscopic resection	1 (1.3)

EMR: Endoscopic submucosal resection; ESD: Endoscopic submucosal dissection; LGD: Low-grade dysplasia; HGD: High grade dysplasia.

Active H. pylori infection was histologically diagnosed in four patients (5.1%), 56 patients (70.9%) were not infected, and 18 patients (22.8%) had a history of eradicated H. pylori infection. Histopathologic examination of the underlying gastric mucosa revealed normal mucosa in 11 patients (13.9%), whereas atrophic gastritis and intestinal metaplasia were observed in 49.4% and 36.7% of the patients, respectively.

Histopathologic examination of the lesions revealed intestinal adenoma in 49 (62.0%) lesions, foveolar adenoma in 23 (29.1%) lesions, and pyloric adenoma in 6 (7.6%) lesions, whereas only 1 (1.3%) lesion was an oxyntic adenoma. Twenty-two (27.8%) lesions harbored low-grade dysplasia (LGD), 18 (22.8%) had HGD, 26 (26.6%) had intramucosal (T1a) adenocarcinoma, and 18 (22.8%) had submucosal (T1b) carcinoma (Table 2).

Clinical and endoscopic outcomes

Patients were followed for an average of 26.3 ± 21.8 months. Residual adenoma, detected within six months of resection, occurred in four patients (5.1%). These patients had initial HGD, with two undergoing piecemeal EMR and two undergoing complete (R0) ESD. Local recurrence was observed in nine patients (11.4%), with only one patient (1.3%) experiencing a metachronous recurrence. The average time to recurrence was 18.0 ± 15.4 months.

Among the 10 patients with recurrent lesions, three patients experienced more than two recurrences. Histological analysis of the recurrent lesions revealed LGD in 8.9%, HGD in 3%, and intramucosal adenocarcinoma in 1.3% of the patients (Table 3).

Table 3 Clinical and endoscopic outcomes (n = 79).

	n (%)
Mean follow-up, months, mean ± SD	26.3 ± 21.8
Early follow-up (3-6 months)
Residual adenoma	4 (5.1)
Management of residual adenoma
EMR	2 (2.5)
ESD	2 (2.5)
Late follow up (> 6 months)
Mean time to recurrence ± SD, months	18.0 ± 15
Recurrent adenoma
Local	9 (11.4)
Metachronous	1 (1.3)
Management of recurrent adenoma
EMR	6 (7.6)
ESD	4 (5.1)
Pathology of recurrent adenoma
LGD	7 (8.9)
HGD	2 (2.5)
Intramucosal adenocarcinoma (T1a)	1 (1.3)

EMR: Endoscopic submucosal resection; ESD: Endoscopic submucosal dissection; LGD: Low-grade dysplasia; HGD: High grade dysplasia.

Gastric adenomas

The sex, mean BMI and comorbidities were comparable among the three groups, and the mean age was significantly greater in the intestinal group than in the other two groups (71.1 ± 12.9 years vs 63.8 ± 15.7 years vs 58.2 ± 17.9 years, P = 0.031 for the intestinal, foveolar and pyloric adenoma groups, respectively). Intestinal adenomas were significantly associated with a family history of gastric adenocarcinoma (P = 0.006) and a history of statin intake (P = 0.040) compared with the other two groups. Foveolar adenomas were significantly associated with a history of NSAID/aspirin intake compared with the other two groups (P = 0.040).

There was no significant difference across the three subgroup groups in either personal history of gastric neoplasia (P = 0.106) or hereditary predisposition to gastric neoplasia (P = 0.106) (Table 4).

Table 4 Clinical, endoscopic and histopathological characteristics of different gastric adenoma subtypes, mean ± SD.

	Intestinal (n = 49) (%)	Foveolar (n = 23) (%)	Pyloric (n = 6) (%)	P value
Patients characteristics:
Gender (male)	69.40	65.20	33	0.215
Mean age, years	71.1 ± 12.9	63.8 ± 15.7	58.2 ± 17.9	0.031
Mean BMI, kg/m2	25.3 ± 3.2	24.1 ± 4.8	25.2 ± 6.2	0.564
Comorbidities
Diabetes	22.40	21.70	16.70	0.94
Cardiovascular diseases	55.10	73.90	33.30	0.142
Morbid obesity	6.10	0.00	0.00	0.389
Malignancy	24.50	30.40	50.00	0.43
Chronic kidney disease	6.10	4.30	0.00	0.79
Cirrhosis	2.00	0.00	16.70	0.069
History of PPI intake	91.80	100	100.00	0.531
History of NSAID/Aspirin	65.30	82.60	33.30	0.046
History of statin	38.80	26.10	0.00	0.04
Hereditary predisposition
FAP	2.00	8.70	16.70	0.106
Juvenile polyposis	0.00	8.70	0.00
Personal history				0.106
Gastric neoplasia
Adenocarcinoma	22.40	4.30	33.30
Adenoma	4.10	13	33.30
Family history				0.006
Gastric adenocarcinoma	14.30	4.30	0
Colorectal cancer	6.10	8.70	0
Other cancers	0	4.30	33.30
Baseline endoscopic findings:
Number of lesions				< 0.001
1/2/≥ 3	93.9/6.1/0	78.3/0/21.7	66.7/33.3/ 0
Mean size, mm	27.5 ± 14.6	34.6 ± 18.1	59.2 ± 33.2	< 0.001
Location				0.048
Cardia/Fundus	24.50	30.40	33.30
Greater curvature	14.30	34.80	33.30
Lesser curvature	8.20	17.40	0.00
Antrum	48.90	17.40	16.70
Pylorus	4.10	0.00	16.70
Gross morphology				0.046
(Paris classification)
0-Ip	2.00	26.10	33.30
0-Is	18.40	21.70	50.00
0-IIa	32.70	21.70	16.70
0-IIb	4.10	0.00	0.00
0-IIc	10.20	4.30	0.00
0-IIa+IIc	28.60	26.10	0.00
0-Is+IIc	4.10	0.00	0.00
Color				0.269
Same as background mucosa	32.70	17.40	0.00
Whitish	2.00	4.30	0.00
Reddish	42.90	60.90	83.30
Ulcerated/Erosive	22.40	17.40	16.70
Treatments
EMR/ESD	6.1/93.9	34.8/65.2	33.3/66.7	0.005
Histological characteristics:
Active Helicobacter pylori	4.10	8.70	0.00	0.373
Underlying stomach
Atrophic gastritis	67.30	13	50	< 0.001
Intestinal metaplasia	26.50	69.60	0.00
Worst histology				0.411
LGD	20.40	30.40	50
HGD	18.40	26.10	33.30
Early gastric adenocarcinoma	57.10	43.50	16.70
T1a	34.70	17.40	0.00
T1b	22.40	26.10	16.70
Mean size of resected specimen, mm	34.1 ± 14.0	47.0 ± 58.0	35.0 ± 23.0	0.349
Histologically complete resection (R0)	89.80	91.30	100	0.935
Clinical outcomes:
Mean follow-up, months	24.9 ± 21.9	26.3 ± 21.4	43.1 ± 27.1	0.613
Early follow-up
Residual adenoma	2.00	13.00	0.00	0.184
Management: EMR/ESD	0.0/2.0	8.7/4.3	0.0/0.0	0.205
Late follow up (recurrence)
Mean time to recurrence, months	21.4 ± 16.8	19.9 ± 17.3	14.3	0.833
Local/metachronous	6.1/2.0	17.4/0.0	16.7/0.0	0.07
Management: EMR/ESD	2.0/6	17.4/0.0	0.0/16.7	0.044
Pathology				0.284
LGD	2.00	17.40	16.70
HGD	4.10	0.00	0.00
T1a	2.00	0.00	0.00

BMI: Body mass index; NSAID: Non steroidal anti-inflamatory drugs; PPI: Proton puump inhibitors; FAP: Familial adenomatous polyposis; EMR: Endoscopic mucosal resection; ESD: Endoscopic submucosal dissection; LGD: Low grade dysplesia; HGD: High grade dysplesia.

With respect to the baseline endoscopic findings, foveolar adenomas were significantly associated with multiple (≥ 3) lesions (P < 0.001) and Paris 0-IIa+IIc classification (P = 0.046). Moreover, the mean size of the largest polyp was significantly greater in the pyloric group (P < 0.001).

Compared with the other two groups, the histological examination of the surrounding mucosa revealed more atrophic gastritis in the intestinal adenoma group (33/49 patients, 67%), whereas the foveolar adenoma group was more frequently associated with intestinal metaplasia (16/23 patients, 70%) (P < 0.001) (Table 4).

ESD was significantly more frequently used in the intestinal adenoma group (94% vs 65% vs 67%, P = 0.005, for the intestinal, foveolar and pyloric adenoma groups, respectively), with variable rates of histologically complete (R0) resection (90% vs 29% vs 100%, P = 0.935, for the intestinal, foveolar and pyloric adenoma groups, respectively).

Early gastric adenocarcinoma (T1a and T1b) was observed in 28/49 (57.1%) of the intestinal adenomas, 10/23 (43.5%) of the foveolar adenomas, and 1/6 (16.7%) of the pyloric adenomas (P = 0.411) (Table 4).

The post ER outcomes were comparable among the three groups, with overall recurrence rates of 8.1% for intestinal adenomas, 17.4% for foveolar adenomas, and 16.7% for pyloric adenomas (P = 0.070). The mean time to recurrence among the three groups was also comparable (21.4 ± 16.8 months vs 19.9 ± 17.3 months vs 14.3 months, P = 0.833, for the intestinal, foveolar and pyloric adenoma groups, respectively) (Table 4).

Ten patients (9 patients with intestinal adenoma and 1 patient with foveolar adenoma) had infiltrating adenocarcinoma and were referred for post-ESD surgery; 6 (7.6%) patients underwent total gastrectomy, 3 (3.8%) underwent partial gastrectomy, and 1 (1.3%) underwent Ivor Lewis esophagectomy.

Potential risk factors for different gastric adenomas

Intestinal adenoma: Age ≥ 65 years (OR = 3.224; 95%CI: 1.198-4.680), a history of statin intake (OR = 1.161; 95%CI: 1.161-3.255), and a family history of gastric adenocarcinoma (OR = 1.641; 95%CI: 1.38-3.995) were identified as potential risk factors associated with intestinal adenoma. In addition to atrophic gastritis (OR 1.831; 95%CI: 1.560-4.329), a single lesion (OR = 1.736; 95%CI: 1.470-4.163), an antral location of the lesion (OR = 2.366; 95%CI: 2.06-5.235) and a Paris 0-IIa morphology (OR = 2.134; 95%CI: 1.840-4.848) were detected (Table 5; Figure 2A-C).

Figure 2 White-light endoscopic and H&E-stained microscopy images of different subtypes of gastric adenomas. A: Endoscopic image (intestinal adenoma); B: Intestinal metaplasia and intestinal adenoma with intramucosal carcinoma (intestinal adenoma); C: Chronic atrophic gastritis with intestinal metaplasia (intestinal adenoma); D: Endoscopic image (foveolar adenoma); E: Foveolar adenoma; F: Chronic gastritis with foveolar hyperplasia (foveolar adenoma); G: Endoscopic image (pyloric adenoma); H: Pyloric adenoma; I: Low-grade dysplasia with intestinal metaplasia (pyloric adenoma).

Table 5 Potential risk factors for different subtypes of gastric adenomas and independent predictors for recurrence after endoscopic resection.

	B-estimate	P value	Odds ratio	95%CI
Potential risk factors for different subtypes of gastric adenomas
Intestinal adenoma:
Age ≥ 65 years	1.171	0.021	3.224	(1.198-4.680)
History of statin intake	0.149	< 0.001	1.161	(1.1-3.255)
Family history of gastric adenocarcinoma	0.495	< 0.001	1.641	(1.38-3.995)
Single polyp	0.551	< 0.001	1.736	(1.47-4.163)
Antrum as location	0.861	< 0.001	2.366	(2.06-5.235)
Atrophic gastritis	0.605	< 0.001	1.831	(1.56-4.329)
Morphology as Paris classification 0-IIa	0.758	< 0.001	2.134	(1.84-4.848)
Foveolar adenoma:
BMI ≥ 28 kg/m2	0.105	< 0.001	1.111	(1.03-3.656)
History of NSAID intake	0.623	< 0.001	1.864	(1.59-4.387)
Morphology as Paris classification 0-IIa	0.927	0.048	2.528	(2.21-5.501)
Intestinal metaplasia	0.513	< 0.001	1.671	(1.41-4.048)
Pyloric adenoma:
BMI ≥ 28 kg/m2	0.442	< 0.001	1.560	(1.26-6.152)
Largest polyp size ≥ 30 mm	0.712	< 0.001	2.039	(1.34-5.11)
Multiple polyps	0.870	< 0.001	2.386	(2.08-5.268)
Morphology as Paris classification 0-Ip	0.863	0.001	2.370	(2.06-5.242)
Independent predictors for recurrence after endoscopic resection
Family history of gastric adenocarcinoma	0.287	< 0.001	1.332	(1.1-3.431)
Largest polyp size ≥ 30 mm	0.714	< 0.001	2.042	(1.76-4.692)
Number of polyps > 3	0.411	< 0.001	1.509	(1.26-3.757)
Lesser curvature as location	0.542	< 0.001	1.720	(1.46-4.135)
Morphology as Paris classification 0-IIc	0.127	< 0.001	1.135	(1.092-3.055)
Active Helicobacter pylori infection	0.139	< 0.001	1.149	(1.019-3.083)

BMI: Body mass index; NSAID: Non steroidal anti-inflamatory drugs.

Foveolar adenoma: BMI ≥ 28 kg/m² (OR 1.111; 95%CI: 1.030-3.656), history of NSAID intake (OR 1.864; 95%CI: 1.590-4.387), largest polyp size ≥ 30 mm (OR 1.560; 95%CI: 1.060-4.160), Paris classification 0-IIa (OR 2.528; 95%CI: 2.210-5.501) and underlying intestinal metaplasia (OR 1.671; 95%CI: 1.410-4.048) were all potential risk factors associated with foveolar adenoma (Table 5; Figure 2D-F).

Pyloric adenoma: BMI ≥ 28 kg/m² (OR 1.560; 95%CI: 1.260-6.152), largest polyp size ≥ 30 mm (OR 2.039; 95%CI: 1.340-5.110), multiple polyps (OR 2.386; 95%CI: 2.08-5.268) and Paris classification 0-Ip (OR 2.370; 95%CI: 2.060-5.242) were all potential risk factors associated with pyloric adenoma (Table 5; Figure 2G-I).

Oxyntic adenoma

The only patient with oxyntic adenoma was an 80-year-old female patient with a history of cardiovascular morbidity and a BMI of 25 kg/m². She had a history of PPI, statin, and NSAID use and no personal or family history of cancer. She complained of epigastric pain, and endoscopy revealed 10 polyps at the greater curvature; the largest polyp diameter was 25 mm. The polyps were completely removed via EMR, revealing oxyntic adenoma with LGD with normal underlying gastric mucosa and no active H. pylori infection. A 3-month follow-up endoscopy revealed a residual LGD adenoma resected via EMR.

Gastric adenoma subtypes and post-ER recurrence

A higher post-ER recurrence rate was observed in the foveolar group (20% at 1 year, 28% at 2 years) than in the intestinal group (5.9% at 1 year, 14% at 2 years) and the pyloric group (0% at 1 year, 20% at 2 years); however, the difference was not statistically significant (log rank P = 0.19) (Figure 3).

Figure 3 Postendoscopic resection recurrence rates according to each subtype of gastric adenoma. A high postendoscopic resection recurrence rate was observed in the foveolar group (20% at 1 year, 28% at 2 years) compared with the pyloric group (0% at 1 year, 20% at 2 years) and the intestinal group (0% at 1 year, 14% at 2 years); however, the difference was not statistically significant (log rank P = 0.19). ER: Endoscopic resection.

Multivariate Cox regression analysis of the independent predictors for adenoma recurrence after ER revealed that a family history of gastric adenocarcinoma (HR = 1.332; 95%CI: 1.10-3.431, P < 0.001) and active H. pylori infection (HR = 1.149; 95%CI: 1.019-3.083, P < 0.001) were independent predictors of recurrence after ER in different gastric adenomas. Among the baseline endoscopic characteristics, a largest polyp size 30 mm (HR = 2.042; 95%CI: 1.760-4.692, P < 0.001), more than 3 polyps (HR = 1.509; 95%CI: 1.260-3.757, P < 0.001), a lesser curvature as the lesion location (HR = 1.720, 95%CI: 1.460-4.135, P < 0.001) and a Paris classification of 0-IIc (HR = 1.135, 95%CI: 1.092-3.055, P < 0.001) were independent predictors of post-ER recurrence (Table 5).

DISCUSSION

Understanding the distinct features of gastric adenoma subtypes is crucial for appropriate management. A retrospective analysis of 79 endoscopically resected lesions was performed and revealed that 38% of the adenomas were of the non-intestinal type. Intestinal adenomas were more likely to occur in older patients (≥ 65) with a history of gastric adenocarcinoma and atrophic gastritis. Foveolar adenomas, on the other hand, displayed associations with NSAID exposure, BMI ≥ 28 kg/m², Paris 0-IIa morphology, and intestinal metaplasia. Pyloric adenomas were characterized by a higher prevalence in patients with BMI ≥ 28 kg/m², large (≥ 30 mm) multiple polyps, and Paris Ip morphology.

Gastric adenoma subtypes display heterogeneity, with intestinal types being the most prevalent. Interpretation of data concerning less common subtypes should consider this context. This study includes a limited number of non-intestinal adenomas, specifically six pyloric adenomas and one oxyntic adenoma, which constrains comprehensive subtype comparisons and may influence the applicability of results to these rarer subtypes.

Prior studies report gastric adenoma recurrence rates after ER ranging from 3%-20.9%[11,18]. In our study, with a 26.3 ± 21.8-month follow-up, residual adenoma occurred in 5.1% and recurrence in 12.7% of patients. Foveolar and pyloric adenomas showed higher recurrence rates than intestinal adenomas (17.4%, 16.7%, and 8.1%, respectively, P = 0.070), although mean recurrence times were similar (P = 0.833). Consistent with previous findings, advanced histology did not predict post-ER recurrence in patients with gastric adenomas[11,19].

A significant 43.5% of foveolar adenomas in our study were found in conjunction with early gastric adenocarcinoma, highlighting their potential for malignant transformation. This observation reinforces existing research that advocates for complete en bloc resection and meticulous follow-up to accurately distinguish foveolar adenomas from hyperplasia and manage their inherent risk[19-21].

The increased recurrence tendency of pyloric adenomas necessitates vigilant follow-up. While infrequent, these lesions are clinically significant due to their malignant potential and the substantial association with adenocarcinoma, reported between 12% and 30% in large-scale studies[22,23].

The present investigation identified a family history of gastric adenocarcinoma, active H. pylori infection, polyp size ≥ 30 mm, multiple polyps (> 3), and Paris 0-IIc morphology as independent predictive factors for post-ER recurrence of gastric adenomas. This highlights the importance of considering these factors during patient management and follow-up.

Furthermore, studies consistently indicate that larger early gastric adenocarcinoma (> 2 cm) are more prone to recurrence[2,11,24]. Despite some variability in published research[25], our study identified active H. pylori infection as an independent predictor of post-ER recurrence. This observation is consistent with other study demonstrating the beneficial impact of H. pylori eradication in mitigating metachronous neoplastic recurrence[26].

The findings of our study have direct clinical implications for patient management and surveillance strategies with some concluded recommendations: (1) Implement risk-stratified surveillance intervals with 3-months follow up endoscopy: Patients with high-risk features (family history, large polyps ≥ 30 mm, > 3 polyps, or Paris 0-IIc morphology), while those without these risk factors could potentially be surveilled at 6-month intervals; (2) Prioritize H. pylori testing and eradication in all gastric adenoma patients to address this modifiable risk factor; (3) Consider more aggressive endoscopic approaches for pyloric adenomas, given that they are associated with larger size and multiplicity; and (4) Maintain heightened surveillance for foveolar adenomas due to their higher recurrence rates and association with early gastric adenocarcinoma.

To our knowledge, this is the first Western study to provide a detailed characterization of gastric adenoma subtypes, bridging a significant gap in literature. The prospective, single-center design, coupled with long-term follow-up and expert pathological review of all ER cases, including hyperplastic lesions, lends credibility to our findings. However, limitations such as the small sample size of non-intestinal adenomas and heterogeneous resection techniques necessitate further validation through larger, multi-center studies. Future research should focus on elucidating the underlying mechanisms driving subtype-specific development and recurrence, particularly for oxyntic adenomas.

CONCLUSION

This study identifies distinct risk factor profiles for different subtypes of gastric adenomas, suggesting that intestinal, foveolar, and pyloric adenomas may have different underlying pathogenic mechanisms. Intestinal adenomas were significantly associated with age ≥ 65 years, statin use, family history of gastric adenocarcinoma, atrophic gastritis, and antral location, while foveolar adenomas were linked to higher BMI, NSAID use, and intestinal metaplasia. Pyloric adenomas were more likely to present as large or multiple polyps, particularly in patients with higher BMI. Furthermore, several independent predictors of recurrence after ER were identified, including family history of gastric neoplasia, largest polyp size ≥ 30 mm, presence of more than three polyps, lesion location on the lesser curvature, Paris 0-IIc morphology, and active H. pylori infection. These findings underscore the importance of subtype-specific risk assessment and tailored surveillance strategies to improve clinical outcomes and reduce recurrence rates in patients undergoing ER for gastric adenomas.