Gastric cancer (GC) is one of the most common malignant tumors of the digestive system worldwide [1]. Benefiting from the improvement and popularization of endoscopic technology in recent years, researchers have discovered many low-grade well differentiated gastric tumors, such as the gastric epithelial neoplasm of the fundic-gland mucosa lineages (GEN-FGML). GEN-FGMLs are rare tumors that differentiate from gastric mucosa to fundic glands, including oxyntic gland adenoma (OGA), gastric adenocarcinoma of the fundic-gland type (GA-FG), and gastric adenocarcinoma of the fundic-gland mucosa type (GA-FGM) [2–4]. OGA and GA-FG are defined as well-differentiated neoplasms composed of parietal cell-like tumor cells with mild morphologic atypia that are positive for pepsinogen I and/or H⁺/K⁺-ATPase. GA-FG can be divided into three subcategories according to the tumor composition: chief cell predominant, parietal cell predominant, and mixed phenotype [2]. The histologic features of GA-FGM are similar to those of GA-FG, but the gastric pit epithelium is atypical and malignant [5, 6]. GA-FGM was also divided into three subtypes (Type 1, Type 2, and Type 3) according to the mucosal structure of the pit epithelium and fundic gland [4]. GEN-FGML can be easily misdiagnosed. GA-FG is known as an oxyntic mucosal polyp/adenoma in the West [7], whereas Japanese scholars believe that OGA is the intramucosal stage of GA-FG. Japanese scholars divided fundic gland tumors into GA-FG and GA-FGM, and GA-FGM is believed to have more malignant potential [6, 8]. In addition, OGA and GA-FG were officially listed as new gastric neoplasms in the World Health Organization (WHO) classification of tumors in 2019 [9]. GA-FG is diagnosed when OGA invades the submucosal layer. GEN-FGMLs, particularly OGA and GA-FG, have a good prognosis [2]. Unlike other low-grade well differentiated gastric tumors, GEN-FGMLs have a high frequency of GNAS mutations, which are a characteristic genetic feature of GEN-FGMLs [4]. Here, we report the clinicopathological features of 37 cases of GEN-FGMLs, helping pathologists and endoscopists diagnose GEN-FGMLs.

Recently, the incidence of early gastric cancer has increased significantly with the application of narrow-band imaging (NBI) magnifying endoscopy [10]. Many H. Pylori-negative gastric adenocarcinomas, including GEN-FGML, have been reported [2, 6, 10–17]. In 2021, GEN-FGML was classified into OGA, GA-FG, and GA-FGM, based on cell differentiation and histological features [4]. Furthermore, GA-FG can be divided into three subcategories according to the tumor composition: chief cell predominant, parietal cell predominant, and mixed phenotype [9]. And GA-FGM was also divided into three subtypes (Type 1, Type 2, and Type 3) according to the mucosal structure of the pit epithelium and fundic gland [4].

In our study, we present 37 cases of GEN-FGMLs, including 19 cases of OGA, 16 cases of GA-FG, and two cases of GA-FGM. Previous studies in Japan have shown that GEN-FGMLs are predominantly observed in males [4]. However, our results showed that GEN-FGMLs are more common in females in China. In the present study, GEN-FGMLs were different based on their sizes. GA-FG and GA-FGM had larger diameters, with an average of 8.8 mm, than OGA, with an average diameter of 5 mm. Furthermore, our study demonstrated that GA-FG and GA-FGM had proliferation indices with Ki-67 than OGA, suggesting that Ki-67 plays a crucial role in the diagnosis of GEN-FGML. Previous studies showed that the background mucosa of GEN-FGML shows no atrophic changes and intestinal metaplasia [4, 16]. In the present study, over 40% of patients had slight atrophic gastritis in the adjacent mucosa.

Histopathological assessment is essential for confirming the diagnosis of GEN-FGMLs. According to the WHO classification, the depth of invasion differentiates between OGA and GA-FG [9]. Based on the Japanese diagnostic criteria, despite showing low-grade atypia and invading the submucosal layer, many stomach neoplasms are diagnosed as invasive carcinomas in the absence of an obvious desmoplastic response [2, 18, 19]. Thus, Japanese scholars have proposed that OGA is the intramucosal stage of GA-FG, and OGA and GA-FG represent different developmental stages of the same tumor [8]. The identification of OGA and GA-FG as intramucosal adenomas or low-grade gastric adenocarcinomas relies on whether the pathologist uses the Western or Japanese diagnostic criteria. In this cohort study, our diagnosis followed the WHO criteria, as we believe that overdiagnosis may cause unnecessary overtreatment in patients in China.

OGA and GA-FG are well-differentiated tumors with expansive and dense growth, composed of pale gray-blue basophilic columnar cells with mild nuclear atypia and slightly enlarged fundic gland cells. IHC staining revealed that they were positive for pepsinogen-I. The surface of the lesions was covered with a normal gastric pit epithelium. In addition, the deep part of the tumor was irregularly branched or dendritic. The tumor gland and background mucosa were demarcated without migrating. The histological features of GA-FGM were similar to those of GA-FG. However, GA-FGM is always accompanied by differentiation to the fundic glands and pit mucous cells and occasionally progresses into tumors consisting of pyloric glands. Differentiation into fundic glands and pit mucous cells suggests that GA-FGM presumably possesses multilineage differentiation potential. GA-FGM is a unique type of complex gastric adenocarcinoma with differentiated fundic glands [2]. Our study revealed histological characteristics similar to those of GEN-FGMLs. Compared to OGA and GA-FG, GA-FGM exhibits greater cellular atypia and malignant transformation of the pit epithelium [6]. GA-FGM showed the highest malignant potential due to higher rates of vascular invasion and lymph node metastasis [2]. Owing to the limited sample size, we did not observe lymphatic invasion and vascular invasion in GA-FGM. However, lymph node metastasis was observed in one GA-FG case. The IHC staining results were consistent with those of previous studies. In the gastric mucosa, pit mucous cells express MUC5AC, gland mucous cells express MUC6, chief cells express pepsinogen I, and parietal cells express H⁺/K⁺-ATPase. GEN-FGMLs mainly consist of tumor components with chief cell-like differentiation, whereas gland mucous cells were previously thought to be the precursors of chief cells [20]. All GEN-FGML samples were positive for MUC6 and pepsinogen-I, indicating that the GEN-FGMLs originated from immature chief cells. MUC5AC was expressed in the pit epithelium and GA-FGM surface, whereas MUC2 and CD10 were not detected. GEN-FGMLs also expressed neuroendocrine markers (Syn and CD56) but not CgA. GEN-FGMLs had a low Ki-67/MIB1 proliferation index and lacked p53 protein.

GEN-FGMLs contain GNAS mutation, which is thought to be a characteristic genetic feature [4]. Ueyama et al. [4] showed that GEN-FGMLs belong to the same genetic lineage, and that GA-FGM may be the final stage of OGA and GA-FG. In addition to GNAS mutations, GEN-FGMLs have genetic mutations associated with Wnt/β-catenin, KRAS, PIK3CA, and sonic hedgehog signalling pathways, but not with TP53 [6, 21–25]. Thus, Wnt/β-catenin signalling pathway may play a role in the development and progression of GEN-FGMLs [6], suggesting that targeting the Wnt/β-catenin pathway or GNAS mutations may be a potential therapeutic strategy for treating GEN-FGMLs. Further studies are needed to fully understand the molecular mechanisms underlying GEN-FGML.

Previous studies on the origin of GEN-FGMLs have not been consistent and have suggested that spasmolytic polypeptide-expressing metaplasia (SPEM) is fundamental to the development of intestinal metaplasia and gastric cancer, and that mature chief cells cannot dedifferentiate into SPEM. However, recent findings suggest that chief cells undergo differentiation into SPEM [26]. Leushacker et al. [27] hypothesized that Lgr5⁺ chief cells are the major origin of gastric cancer because Lgr5⁺ cells play an essential role in maintaining homeostatic stem cells. It has been shown that chief cells can dedifferentiate into different types of gastric mucosal epithelium, such as pit mucosal cells and gland cells. These results suggest that GEN-FGMLs may be derived from malignant transformation of the chief cells. Recent studies have indicated that intramucosal gastric carcinoma (IGC), OGA, and SPEM are derived from common ancestral glands. IGC and OGA originate from the KRAS-mutated SPEM [28]. However, the specific mechanism needs to be investigated further.

It is crucial to distinguish GEN-FGMLs from other gastric tumors. MUC6 and pepsinogen-I co-expression are crucial for the diagnosis of GEN-FGMLs, but the use of simple IHC markers may result in misdiagnosis. All participants in this study were positive for CD56 and Syn, and negative for CgA. These results could have led to the misdiagnosis of neuroendocrine tumors. Gastric neuroendocrine tumors are organ-like structures with basophilic tumor cells, accompanied by significant interstitial microvascular hyperplasia and CgA expression compared to GEN-FGMLs, which do not express CgA [29]. Distinguishing between GEN-FGMLs and pyloric gland adenomas (PGAs) is also challenging. Both PGA and GEN-FGMLs develop in the gastric fundus mucosa and express MUC6, pepsinogen-I, and MUC5AC. Nonetheless, PGAs show polypoid proliferation of pyloric-type glands with cuboidal/columnar cells and foamy ground-glass cytoplasm. Kushima et al. revealed that the outer tumor cells in PGAs are taller than the inner cells [25]. Differentiated adenocarcinoma with low atypia was related to GA-FG, but it lacked chief or parietal cells and showed an aggressive growth pattern despite the low atypia. Owing to the rarity of GEN-FGMLs and lack of awareness, most of our cases were primarily misdiagnosed as fundic gland polyps (FGPs). Both FGPs and GEN-FGMLs originate from fundic glands and exhibit similar features, such as the presence of hyperplastic glandular structures, cystic dilation, and distortion of glandular architecture. Typically, FGPs are benign lesions with less than 5 mm in diameter, consisting of dilated oxynitic glands, foveolar hypoplasia, and parietal hyperplasia [9]. However, GEN-FGMLs originate from the fundic glands of the stomach and display glandular or tubular structures with a monoclonal shape. FGPs can also exhibit intraepithelial neoplasia and dysplasia [30]. Notably, dysplasia or adenocarcinoma in fundic gland polyps originates from the foveolar epithelium and not from the fundic glands. It is difficult to diagnose GEN-FGMLs based on biopsy specimens because the mucosal muscle may be indiscernible in such samples. Previous studies and our study have shown that GEN-FGML lesions with larger sizes are more often diagnosed as GA-FG or GA-FGM [4, 16]. Furthermore, our study demonstrated that GA-FG and GA-FGM had higher proliferation indices with Ki-67 than OGA, suggesting that Ki-67 plays a crucial role in the diagnosis of GEN-FGML, especially in biopsy specimens. Specimens with Ki-67 proliferation indices >2.5% and lesion size >4.5 mm are more likely to be diagnosed with GA-FG and GA-FGM than OGA. Moreover, endoscopic findings resemble those of early undifferentiated carcinoma and mucosa-associated lymphoid tissue. Endoscopy plays a limited role in the differential diagnosis. NBI is a promising technique for the diagnosis of GA-FGM and GA-FG. There are distinctive differences between GA-FGM and GA-FG in conventional endoscopy and NBI, including indigo carmine dye spraying, demarcation line, peculiar microvascular pattern, irregular microsurface pattern, and wider marginal crypt epithelium [13].

GEN-FGMLs are lower aggressive than other GCs as they possess minor cellular atypia, less vascular invasion and proliferation, and lack TP53 mutations. GA-FGM, particularly type 2 GA-FGM, displays the highest malignant potential owing to strong vascular invasion, lymph node metastasis, and deep submucosal invasion [4]. Standard indications for endoscopic and surgical treatment of GEN-FGMLs are urgently needed. Two retrospective analyses from Japan have shown that endoscopic resection is currently the most common and appropriate initial treatment for GEN-FGMLs [4, 16]. Iwamuro et al. [16] indicated that GA-FG with an invasion depth of > 500 μm will not recur, even if endoscopic resection is the only treatment, as muscle infiltration and vascular invasion are uncommon in GA-FG. Ueyama et al. [4] demonstrated that no cases of GEN-FGMLs (including GA-FGM) that underwent endoscopic resection experienced recurrence, metastasis, or death. The Japanese Gastric Cancer Treatment Guidelines (6th edition) stipulate that radical tumor removal must be performed in differentiated gastric cancer when the invasion depth is > 500 μm [3]. Thus, the treatment of GEN FGML is challenging. On the one hand, surgical resection may not be an advisable treatment for GA-FG with an invasive depth of > 500 μm [16]. In contrast, GA-FGM was more aggressive than GA-FG. Therefore, endoscopic treatment guidelines for common gastric adenocarcinomas should be followed to treat GA-FGM. Hence, future studies on GEN-FGMLs should establish standardized endoscopic treatment and follow-up protocols.

In conclusion, GEN-FGML is a rare type of well-differentiated tumors possibly originating from dedifferentiated chief cells, with typical histopathological and molecular features. GEN-FGMLs are a group of well-differentiated gastric tumors with favorable biological behaviors, low cellular atypia, and low proliferation. It can be divided into OGA, GA-FG, and GA-FGM subtypes. Immunohistochemistry is critical for confirming diagnosis. The Ki-67 proliferation indices and lesion size may help differentiate between OGA, GA-FG, and GA-FGM, especially in biopsy specimens. Close cooperation between pathologists and endoscopists is required to avoid misdiagnosis and overtreatment. Endoscopic resection is currently the mainstream initial treatment for GEN-FGMLs.