GNAS mutation as an alternative mechanism of activation of the Wnt/β-catenin

signaling pathway in gastric adenocarcinoma of the fundic gland type

Ryosuke Nomura MDa, b, Tsuyoshi Saito MD, PhDa, Hiroyuki Mitomi MD, PhDc, Yasuhiro Hidaka

MD, PhDa, b, Se-yong Lee MDa, b, Sumio Watanabe MD, PhDb, Takashi Yao MD, PhDa.

a Department of Human Pathology, Juntendo University, School of Medicine, 2-1-1 Hongo,

Bunkyo-ku, Tokyo 113-8421, Japan.

b Department of Gastroenterology, Juntendo University School of Medicine

c Department of Surgical and Molecular Pathology, Dokkyo University School of Medicine,

Tochigi, Japan

Corresponding Author

Tsuyoshi Saito, MD, PhD.

E-mail: tysaitou@juntendo.ac.jp

Department of Human Pathology, Juntendo University, School of Medicine, 2-1-1 Hongo,

Bunkyo-ku, Tokyo 113-8421

Keywords: Gastric adenocarcinoma of the fundic gland type; CTNNB1; Axin; APC; GNAS;

KRAS; mutation

Running title: GNAS mutation in gastric adenocarcinoma of the fundic gland type

Abstract

Gastric adenocarcinoma of the fundic gland type (GAFG) is a rare variant of gastric tumor.

We have recently reported the frequent accumulation of β-catenin in GAFGs and showed

that approximately half of the cases studied harbored at least one mutation in

CTNNB1/AXINs/APC, leading to the constitutive activation of the Wnt/β-catenin pathway.

However, the mechanisms of Wnt signaling activation in the remaining cases are unknown.

Accumulating evidence showed that the activating mutation in GNAS promotes

tumorigenesis via the activation of the Wnt/β-catenin pathway or the ERK1/2 MAPK

pathway. Therefore, we analyzed the mutations in GNAS (exons 8 and 9) and in KRAS

(exon 2) in 26 GAFGs. Immunohistochemistry revealed nuclear β-catenin expression in 22

out of 26 GAFGs, and ten out of 26 cases (38.5%) harbored at least one mutation in

CTNNB1/AXINs/APC. Activating mutations in GNAS were found in five out of 26 GAFGs

(19.2%), all of which harbored R201C mutations. Activating mutations in KRAS were found

in two of 26 GAFGs (7.7%), and both of these also contained GNAS activating mutations.

Four out of five cases with GNAS mutation showed nuclear β-catenin expression, and

presence of GNAS mutation was associated with β-catenin nuclear expression (P=0.01).

Furthermore, three of these four cases did not harbor mutations in CTNNB1, APC, or

AXINs, suggesting that mutations in the Wnt component genes and those in GNAS occur

almost exclusively. These results suggest that GNAS mutation might occur in a small subset

of GAFG as an alternative mechanism of activating the Wnt/β-catenin signaling pathway.

Introduction

Gastric adenocarcinoma of the fundic gland type (GAFG) is a recently described

histological variant of gastric cancer [1-4]. Tumors of this type usually present as solitary,

small (typically, less than 10 mm), and well-circumscribed lesions extending into the

submucosa in the upper third of the stomach. Histologically, they are composed of tightly

packed glands and cords of, predominantly, chief cells with slight nuclear pleomorphism,

lacking chronic gastritis, atrophic change, or intestinal metaplasia in their surrounding

mucosa. Based on the common occurrence of GAFG and the sporadic fundic gland polyp

with frequent CTNNB1 mutation in the fundic gland mucosa, the constitutive activation of

the Wnt signaling pathway is hypothesized in this tumor. However, a previous study

demonstrated a lack of nuclear staining of β-catenin in this tumor [3]. Furthermore, we

have recently shown that there is frequent β-catenin accumulation in GAFG and that

approximately half of these tumors harbored mutations in CTNNB1, APC or AXINs 1 and 2,

leading to the constitutive activation of the Wnt/β-catenin pathway [5]. However, the

mechanism of cancer pathogenesis remains unknown in the remaining cases.

GNAS encodes the stimulatory G-protein alpha subunit (Gsα), which is situated on the

human chromosome 20q13.3. Heterotrimeric G-proteins composed of α-, β-, and γ-subunits

mediate signal transduction from a large number of hormone- and growth factor-activated

seven-transmembrane receptors to control diverse intracellular signaling pathways [6].

Ligand-bound G-protein-coupled receptors activate the Gs-protein by promoting the

exchange of guanosine diphosphate (GDP) for guanosine triphosphate (GTP) on Gsα, which

results in the dissociation of the Gsα subunit from the βγ-complex of the receptor. The

free Gsα subunit interacts with adenylate cyclase to stimulate the synthesis of cyclic

adenosine monophosphate (cAMP) before GTP is hydrolyzed to GDP, causing this inactive

molecule to re-associate with the βγ-complex [6]. Activating mutations of GNAS are

prevalent in benign tumors and in tumors with low-grade malignancies, including pituitary

adenomas [7, 8], intraductal papillary mucinous tumors (IPMTs) of the pancreas [9-11],

villous adenoma of the colorectum [12], pyloric gland adenoma of the stomach [13], and

parosteal osteosarcoma [14]. Conversely, these mutations are absent or rare in most of

other definite malignancies such as conventional adenocarcinoma of the stomach [13]. A

recent study reported that two out of three GAFG cases also harbored a GNAS mutation

[15]; however, this did not fully explain the frequency of activating GNAS mutations in this

tumor [15]. While some studies have reported that GNAS mutations are often associated

with KRAS or BRAF mutation [9, 10, 12, 13], another study demonstrated that the Wnt

signaling pathway could be activated directly or indirectly by GNAS mutation [16].

In the present study, we examined the mutation status of GNAS and KRAS in GAFG to

elucidate how these mutations are associated with activation of the Wnt signaling pathway.

Materials and methods

Tissue samples

Tissue samples were collected from GAFG patients (n = 26) who underwent endoscopic (n

= 19) or surgical (n = 7) resection at Juntendo University Hospital or affiliated hospitals

(Table 1). These patients comprised of 20 from our previous study [5] and six newly

sampled patients. Seven out of the 27 cases analyzed in our previous study [5] were

excluded from the present analysis because of sample shortage. All cases of GAFG were

found to extend into the submucosa. For comparison, samples (n =32) from patients with

sporadic fundic gland polyps (FGPs) were selected from pathological files at Juntendo

University Hospital and affiliated hospitals (Table 2). Sporadic FGP is a neoplastic lesion

arising from fundic gland mucosa, and Wnt signaling is activated by a β-catenin mutation in

this lesion [17-21]. This study was approved by the institutional review board and the

ethical committee of our hospital (registration no. 2012008).

Immunohistochemistry for β-catenin

Four micrometer-thick tissue sections were cut from formalin-fixed and paraffin-embedded

(FFPE) blocks and were subjected to immunohistochemistry using a monoclonal antibody

against β-catenin (clone 14, 1:200 dilution; BD Bioscience, San Diego, CA). Antigen

retrieval was performed in Tris-EDTA buffer (pH 6.0) by heating in an autoclave. Sections

were incubated at 4°C overnight with primary antibody. Immunohistochemical staining was

visualized using an Envision Kit (Dako, Glostrup, Denmark) with substrate-chromogen

solution. The percentages of stained nuclei (labeling index; LI) were evaluated in the most

representative areas showing the highest immunoreactivity by counting the number of β-

catenin positive cells out of a minimum of 500 tumor cells. Nuclear β-catenin LI was

classified as follows: 5% or less, negative; more than 5%, positive [5]. Slides were scored

semi-quantitatively by three independent investigators (authors R.N., Y.H., and T.S.)

without prior knowledge of clinicopathological data. Discrepancies were resolved by

reevaluation until a consensus was reached.

Mutation analysis

We looked for mutations in GNAS (exon 8 and 9), KRAS (exon 2), CTNNB1 (exon 3), APC

(exon 15), AXIN1 (exon 5), and AXIN2 (exon1 and 5) in 26 cases of GAFG to elucidate the

relationship between the activation of Wnt signaling and GNAS/KRAS mutation. We used

polymerase chain reaction (PCR) followed by direct sequencing to observe genetic alterations

using pure tumor genome DNA in GAFG. Genomic DNA was extracted from FFPE sections

using a QIAamp DNA FFPE Tissue kit (Qiagen GmbH, Hilden, Germany). Sections were

stained lightly with hematoxylin, areas of tumor were separated by modified microdissection,

and the tissue samples were observed directly under a light microscope. The corresponding

non-tumor genomic DNA was also extracted where possible. Primer sequences used in this

study have been previously described [5, 13]. Purified PCR products were sequenced with

dideoxynucleotides (BigDye Terminator v3.1; Applied Biosystems, Foster City, CA) and

specific primers, purified using a BigDye X Terminator Purification Kit (Applied Biosystems),

and analyzed with a capillary sequencing machine (3730xl Genetic Analyzer; Applied

Biosystems). Mutations were confirmed where the height of the mutated peak reached 20%

of the height of the normal peak, as previously described [5]. All mutations were verified by

sequencing both the sense and antisense strands. Mutations were evaluated and confirmed

by three independent investigators (authors R.N., Y.H., and T.S.).

Statistical analysis

All statistical analyses were performed using StatView for Windows Version 5.0 (SAS

Institute Inc., Cary, NC). Continuous data were compared by Mann-Whitney U test.

Categorical analysis of variables was performed using either the X2 test (with Yates

correction) or Fisher’s exact test, as appropriate. A p-value less than 0.05 was considered

statistically significant.

Results

Clinicopathological findings

Clinicopathological findings of GAFGs and FGPs are summarized in Table 2. Most of GAFGs

(88.5%) were located in the upper third of the stomach. In this study, a male

predominance for GAFG was observed, and patients’ ages ranged from 49 to 79 years

(mean age: 66 years). Tumor size of GAFGs ranged from 3 to 85 mm (mean tumor size:

15.3 mm). Macroscopically, 15 cases were the superficial elevated type, five cases were the

superficial depressed type, and six were mixed type. Histologically, all GAFGs were the well-

differentiated type. All GAFG cases invaded the submucosal layer, and the depth of invasion

ranged from 50 to 1500 m (mean depth of invasion: 417 m). Lymphatic and venous

invasions were observed in three and two cases, respectively. None of the lesions have

demonstrated a recurrence of disease or metastasis on follow up.

Immunohistochemistry

Nuclear β-catenin LIs ranged from 0.0% to 62.9% (mean: 28.0%, median: 24%) in

GAFGs. Four out of 26 cases (15.4%) were negative for nuclear β-catenin staining. Among

the 22 cases with nuclear β-catenin staining, nine (34.6%) had LI lower than 30%, and 13

(50.0%) had LI higher than 30%. Various expression patterns of β-catenin with or without

mutations in Wnt genes or GNAS are shown in Fig. 1.

Mutation analyses

First, we evaluated mutations in exons 8 and 9 of GNAS and exon 2 of KRAS in GAFGs and

FGPs. Activating mutations in GNAS were found in five out of 26 GAFGs (19.2%), all of

which harbored the same mutations (exon 8, 601C>T, R201C; Fig. 2). Four out of five

cases with GNAS mutation showed nuclear β-catenin expression, and presence of GNAS

mutation was associated with β-catenin nuclear expression (P=0.01). Activating mutations

in KRAS were found in two of 26 GAFGs (7.7%; Fig. 2), both of which contained GNAS

activating mutations. Conversely, none of the FGPs had either GNAS or KRAS mutations

(0/32; 0%; Table 3).

Next, we further evaluated the mutations in exon 3 of CTNNB1, the mutation cluster

region (exon 15) of APC, AXIN1 (exon 5), and AXIN2 (exon 1 and 5) in six new cases of

GAFGs. In total, activating mutations in CTNNB1 were found in two out of 26 GAFGs

(7.7%). Loss of functional mutations in APC and AXIN1 were also found in two out of 26

GAFGs (7.7%), and one of these cases harbored both mutations. Activating mutations in

AXIN2 were found in six out of 26 GAFGs (23.1%), and two of these cases harbored

mutations in both GNAS and KRAS (Table 3). Totally, ten out of 26 cases (38.5%) harbored

at least one mutation in CTNNB1/AXINs/APC, and thirteen out of 26 cases (50%) harbored

at least one mutation in CTNNB1/AXINs/APC/GNAS.

Clinicopathological relationship with GNAS mutation

Next, we investigated the relationship between clinicopathological characteristics and

GNAS mutation in GAFGs. The tumor samples with a GNAS mutation tended to invade the

submucosa more deeply (p = 0.09) and have increased tumor sizes than those without

GNAS mutation (p = 0.08); however, these differences in invasive ability and tumor size

failed to reach statistical significance (Table 4).

Discussion

Activating mutations of GNAS are reported to be prevalent in cancer, from benign tumors

to tumors with low-grade malignant potential such as pituitary adenomas [7, 8], intraductal

papillary mucinous tumors (IPMTs) of the pancreas [9-11], and parosteal osteosarcoma

[14]. However, GNAS mutations are absent or rare in malignant tumors including

conventional adenocarcinoma of the stomach [13]. A recent study reported that two out of

three GAFG cases also harbored a GNAS mutation [15]; however, it is not enough to

recognize the frequency of activating GNAS mutations in this tumor [15]. In our study,

activating mutations in GNAS were found in five out of 26 (19.2%) cases of GAFGs, all of

which harbored the same mutation (exon 8, 601C>T, R201C). The impact of GNAS

mutations on the clinicopathological features was also assessed, and GNAS mutations

seemed to be associated with deep invasion of the submucosa as well as increased tumor

size, although these failed to reach statistical significance. These findings suggest a

possible oncogenic role of GNAS mutations in promoting tumor progression and invasion in

GAFG. However, a recent paper reported that the intestinal expression of GNAS, activated

by these mutations, and the subsequent increase in activated ERK1/2 within the intestine

were insufficient to induce tumorigenesis [16]. Furthermore, it was demonstrated that the

effects of GNAS R201C on the Wnt signaling pathway were additive to the inactivation of

Apc and that activating mutations of GNAS cooperate with inactivation of APC and are likely

to contribute to colorectal tumorigenesis [16]. Consistent with these findings, we found a

KRAS mutation in two out of 26 (7.7%) GAFG samples (c.35G>A p.G12D and c.38G>A

p.G13D, each), both of which also harbored GNAS activating and Axin2 inactivating

mutations. Further, genetic alterations such as BRAF, NRAS, and EGFR, all of which result in

activation of ERK1/2 MAPK, might coexist in the remaining three cases with GNAS mutation

[14].

Next, we will consider the relationship between Wnt signaling activation and GNAS mutation

in GAFG because in this series of GAFGs, only eight out of 22 cases with nuclear β-catenin

staining harbored mutations of Wnt component genes. Several studies have recently

demonstrated that GNAS mutations contribute to tumorigenesis, both directly and indirectly,

via the Wnt signaling pathway [16]. Furthermore, it has been shown that the threshold of

Wnt pathway activation by constitutively active Gsα is lowered following the loss of APC [22].

Therefore, we examined a possible role of the GNAS mutation on the activation of the Wnt

signal pathway in GAFGs. Four out of five GAFGs (80%) with the GNAS mutation showed

nuclear β-catenin expression, and three of those four cases did not harbor mutations in any

of the Wnt component genes. These results suggest that GNAS mutation occurs in a subset

of GAFG as an alternative mechanism for Wnt/β-catenin signal activation and contributes to

tumorigenesis. However, the mechanisms of Wnt signal activation are unclear in the

remaining 11 cases with nuclear β-catenin expression. It was not clear why two out of four

cases with mutations in either Axin2 or GNAS did not express nuclear β-catenin.

Finally, pathologists may argue that GAFG is not a true carcinoma but oxyntic gland

polyp/adenoma, considering its biological potential [3]. However, apart from these debates,

this study demonstrated that GNAS mutation partly contributes to the activation of Wnt

signaling pathway in this tumor.

In conclusion, this study demonstrates that a small subset of GAFG harbors activating GNAS

and KRAS mutations. Further, GNAS mutation may act as an alternative mechanism of Wnt/β-

catenin signaling activation that contributes to tumorigenesis.

Acknowledgments

The authors thank Dr. Minako Hirahashi (Department of Anatomic Pathology, Pathological

Sciences, Graduate School of Medical Sciences, Kyushu University), Dr. Yumi Oshiro

(Department of Pathology, Matsuyama Red Cross Hospital), Dr. Takehiro Tanaka

(Department of Pathology, Okayama University Graduate School of Medicine), Dr. Yutaka

Nakashima (Division of Pathology, Japanese Red Cross Fukuoka Hospital), Dr. Tetsumi

Yamane (Department of Pathology, Tottori Red Cross Hospital), Dr. Fumiyoshi Fushimi

(Department of Pathology, National Kyushu Cancer Center), Dr. Shinji Kono (Division of

Clinical Pathology, Harasanshin Hospital), Dr. Shuichi Ohara (Department of

Gastroenterology, Tohoku Rosai Hospital), Dr. Koyu Suzuki (Department of Pathology, St

Luke's International Hospital), and Dr. Takeshi Yano (Department of Surgery, Asoka Hospital)

for kindly providing samples and clinical information, Dr. Ayumi Osako (Department of

Gastroenterology, Tottori Seikyo Hospital). We also wish to thank Mrs. Keiko Mitani

(Department of Human Pathology, Juntendo University School of Medicine) for her expert

technical assistance. We also thank the Laboratory of Molecular and Biochemical Research,

Research Support Center, Juntendo University Graduate School of Medicine, Tokyo, Japan,

for technical assistance.

Funding

This work was supported, in part, by a Grant-in-Aid for General Scientific Research from

the Ministry of Education, Science, Sports, and Culture (#26670286 to Tsuyoshi Saito,

#24590429 to Hiroyuki Mitomi and #26460428 to Takashi Yao), Tokyo, Japan.

Conflict of interest

The authors declare that there are NO conflicts of interest to disclose.

Author contribution

RN, YH and SL performed experiments. RN, YH, HM, and TS evaluated the

immunohistochemistry and mutational analysis. RN, YH, HM, SW, TY, and TS wrote the

manuscript. HM and TY diagnosed these cases.

References

[1] Ueyama H, Yao T, Nakashima Y, et al. Gastric adenocarcinoma of fundic gland type

(chief cell predominant type): proposal for a new entity of gastric adenocarcinoma. Am J

Surg Pathol 2010; 34, 609-619.

[2] Tsukamoto T, Yokoi T, Maruta S, et al. Gastric adenocarcinoma with chief cell

differentiation. Pathol Int 2007; 57, 517-522.

[3] Singhi AD, Lazenby AJ, Montgomery EA. Gastric adenocarcinoma with chief cell

differentiation: a proposal for reclassification as oxyntic gland polyp/adenoma. Am J Surg

Pathol 2012; 36, 1030-1035.

[4] Chen WC, Rodriguez-Waitkus PM, Barroso A, Balsaver A, McKechnie JC. A Rare Case of

Gastric Fundic Gland Adenocarcinoma (Chief Cell Predominant Type). J Gastrointest Cancer

2012; 43, 262-265.

[5] Hidaka Y, Mitomi H, Saito T, et al. Alteration in the Wnt/β-catenin signaling pathway in

gastric neoplasias of fundic gland (chief cell predominant) type. Hum Pathol 2013; 44,

2438-2448.

[6] Weinstein LS, Liu J, Sakamoto A, Xie T, Chen M. Minireview: GNAS: normal and

abnormal functions. Endocrinology 2004; 145, 5459-5464.

[7] Landis CA, Masters SB, Spada A, Pace AM, Bourne HR, Vallar L. GTPase inhibiting

mutations activate the alpha chain of Gs and stimulate adenylyl cyclase in human pituitary

tumours. Nature 1989; 340, 692-696.

[8] Lyons J, Landis CA, Harsh G, et al. Two G protein oncogenes in human endocrine

tumors. Science 1990; 249, 655-659.

[9] Wu J, Matthaei H, Maitra A, et al. Recurrent GNAS mutations define an unexpected

pathway for pancreatic cyst development. Sci Transl Med 2011; 3, 92ra66.

[10] Furukawa T, Kuboki Y, Tanji E, et al. Whole-exome sequencing uncovers frequent

GNAS mutations in intraductal papillary mucinous neoplasms of the pancreas. Sci Rep 2011;

1, 161.

[11] Amato E, Molin MD, Mafficini A, et al. Targeted next-generation sequencing of cancer

genes dissects the molecular profiles of intraductal papillary neoplasms of the pancreas. J

Pathol 2014; 233:217-227.

[12] Yamada M, Sekine S, Ogawa R, et al. Frequent activating GNAS mutations in villous

adenoma of the colorectum. J Pathol 2012; 228, 113-118.

[13] Matsubara A, Sekine S, Kushima R, et al. Frequent GNAS and KRAS mutations in

pyloric gland adenoma of the stomach and duodenum. J Pathol 2013; 229, 579-587.

[14] Carter JM, Inwards CY, Jin L, et al. Activating GNAS mutations in parosteal

osteosarcoma. Am J Surg Pathol 2014; 38, 402-409.

[15] Kushima R, Sekine S, Matsubara A, Taniguchi H, Ikegami M, Tsuda H. Gastric

adenocarcinoma of the fundic gland type shares common genetic and phenotypic features

with pyloric gland adenoma. Pathol Int 2013; 63, 318-325.

[16] Wilson CH, McIntyre RE, Arends MJ, Adams DJ. The activating mutation R201C in

GNAS promotes intestinal tumourigenesis in Apc(Min/+) mice through activation of Wnt and

ERK1/2 MAPK pathways. Oncogene 2010; 29, 4567-4575.

[17] Abraham SC, Nobukawa B, Giardiello FM, Hamilton SR, Wu TT. Sporadic fundic gland

polyps: common gastric polyps arising through activating mutations in the beta-catenin

gene. Am J Pathol 2001; 158, 1005-1010.

[18] Sekine S, Shibata T, Yamauchi Y, et al. Beta-catenin mutations in sporadic fundic gland

polyps. Virchows Arch 2002; 440, 381-386.

[19] Jalving M, Koornstra JJ, Boersma-van Ek W, et al. Dysplasia in fundic gland polyps is

associated with nuclear beta-catenin expression and relatively high cell turnover rates.

Scand J Gastroenterol 2003; 38, 916-922.

[20] Abraham SC, Nobukawa B, Giardiello FM, Hamilton SR, Wu TT. Fundic gland polyps in

familial adenomatous polyposis: neoplasms with frequent somatic adenomatous polyposis

coli gene alterations. Am J Pathol 2000; 157, 747-754.

[21] Abraham SC, Park SJ, Mugartegui L, Hamilton SR, Wu TT. Sporadic fundic gland polyps

with epithelial dysplasia : evidence for preferential targeting for mutations in the

adenomatous polyposis coli gene. Am J Pathol 2002; 161, 1735-1742.

[22] Castellone MD, Teramoto H, Williams BO, Druey KM, Gutkind JS. Prostaglandin E2

promotes colon cancer cell growth through a Gs-axin-beta-catenin signaling axis. Science

2005; 310, 1504-1510.

Figure legends

Figure 1. β-catenin immunohistochemical staining in two cases of endoscopically resected

GAFG, and various expression pattern of β-catenin with or without mutations in Wnt genes

or GNAS. A–C: GAFG (Case 10) with GNAS mutations. D-F: GAFG (Case 22) with APC

mutation. A, D: H.E. staining. Tumor cells show diffuse membranous and nuclear expression

of β-catenin (B, E). β-catenin nuclear expression can be observed at the invasive edge of the

tumor (C, F). G: β-catenin expression in GAFG (Case 25) with Axin2 mutation. β-catenin

labeling index in this case was 0%. H: β-catenin expression in GAFG (Case 8) with CTNNB1

mutation. β-catenin labeling index in this case was 18%. I: β-catenin expression in GAFG

(Case 22) with APC mutation. β-catenin labeling index in this case was 48%. J: β-catenin

expression in GAFG (Case 26) without mutations in either Wnt genes or GNAS. β-catenin

labeling index in this case was 1%. K: β-catenin expression in GAFG (Case 7) without

mutations in either Wnt genes or GNAS. β-catenin labeling index in this case was 8%. L: β-

catenin expression in GAFG (Case 19) without mutations in either Wnt genes or GNAS. β-

catenin labeling index in this case was 55%.

A, B, D, E: Original magnification x200. C, F-L: Original magnification: x600

Figure 2: GNAS and KRAS mutations in a case of GAFG (Case 23). A: A substitution of C to T

at c.601 is observed only in tumor-derived DNA, resulting in p.R201C in GNAS. B: A

substitution of G to A at c.38 is observed only in tumor-derived DNA, resulting in p.G13D in

KRAS. All samples were bi-directionally sequenced.

Table 1. Clinicopathologic characteristics of 26 cases of gastric adenocarcinoma of fundic gland type.

1 71/M U 85 superficial elevated 300 1 0 -2 61/M U 9 superficial elevated 200 0 0 -3 74/M M 39 mixed 400 1 0 -4 75/F U 15 superficial elevated 200 0 0 -5 49/M U 12 superficial elevated 500 0 1 -6 69/M M 20 mixed 50 0 0 -7 76/M U 10 superficial depressed 200 0 0 -8 59/M U 9 superficial elevated 50 0 0 -9 71/M U 6 superficial elevated 100 0 0 -10 70/M U 15 mixed 1500 0 0 -11 60/M U 7 superficial elevated 100 0 0 -12 67/M U 3 mixed 100 0 0 -13 68/F U 24 mixed 1200 0 0 -14 57/M U 12 superficial elevated 300 0 0 -15 73/F U 3 superficial elevated 50 0 0 -16 65/M M 12 superficial depressed 600 0 0 -17 63/M U 25 superficial elevated 500 0 0 -18 62/M U 12 superficial depressed 1200 0 1 -19 75/M U 6 superficial elevated 100 0 0 -20 76/M U 5 superficial depressed 300 0 0 -21 55/M U 16 superficial depressed 800 0 0 -22 51/M U 4 superficial elevated 100 0 0 -23 66/M U 18 mixed 1500 0 0 -24 58/M U 7 superficial elevated 100 0 0 -25 79/F U 7 superficial elevated 200 0 0 -26 72/M U 17 superficial elevated 200 1 0 -

Sex; Male (M), Female (F)SM: submucoasal; meta: metastasisly: lymphatic invasion, v: venous invasionLocation; U: upper one third of the stomach (Fundus), M: middle one third of the stomach (Corpus)

Depth of SM

invasion (μ

m)

ly vLymph

node metaCase Age/Sex Location Size(mm) Macroscopic type

Table 2. Summary of clinicopathologic data of GAFG and FGP

Age (mean) 66.0 (49-79) 60.4 (27-87) 0.0767Sex

Male 22 17 0.0132 Female 4 15

Locacion Upper third 23 11 ＜0.0001 Middle third 3 19 Lower third 0 2Size of tumor(mm) 15.3 (3-85) 6.8 (2-20) 0.076Macroscopic typesuperficial elevated 15 32 ＜0.0001

superficial depressed 5 0 mixed 6 0

Histologic typeWell differentiated 26 (100%)

Depth of

submucosal417 (50-1500)

Lymphatic invasion 3 (7.7%)Venous invasion 2 (5.4%)

Variable GAFG (n=26) FGP (n=32) p

Table 3. Immunohistochemical and genetic findings in 26 cases of gastric adenocarcinoma of fundic gland type.

GNAS KRAS CTNNB1 APC Axin 1 Axin 2

1 (+) 11 p.R201C p.G12D (-) (-) (-) p.P27L

2 (+) 15 p.R201C (-) (-) (-) (-) (-)3 (+) 13 (-) (-) (-) (-) (-) (-)4 (+) 20 (-) (-) (-) p.S1411N, p.D1498N p.H531N (-)

5 (+) 19 (-) (-) (-) (-) (-) p.N141D, p.P435S, p.P483S

6 (+) 6 (-) (-) p.A21T (-) (-) (-)7 (+) 8 (-) (-) (-) (-) (-) (-)8 (+) 18 (-) (-) p.E9X (-) (-) (-)9 (+) 23 (-) (-) (-) (-) (-) (-)10 (+) 24 p.R201C (-) (-) (-) (-) (-)11 (+) 56 p.R201C (-) (-) (-) (-) (-)12 (+) 25 (-) (-) (-) (-) p.H531R, p.H540R p.A113V, p.A133V13 (+) 23 (-) (-) (-) (-) (-) (-)14 (+) 46 (-) (-) (-) (-) (-) (-)15 (+) 54 (-) (-) (-) (-) (-) (-)16 (+) 40 (-) (-) (-) (-) (-) p.T153N17 (+) 63 (-) (-) (-) (-) (-) (-)18 (+) 50 (-) (-) (-) (-) (-) (-)19 (+) 55 (-) (-) (-) (-) (-) (-)20 (+) 42 (-) (-) (-) (-) (-) (-)21 (+) 46 (-) (-) (-) (-) (-) (-)22 (+) 48 (-) (-) (-) p.D1484N (-) (-)23 (-) 0 p.R201C p.G13D (-) (-) (-) p.T153N24 (-) 0 (-) (-) (-) (-) (-) (-)25 (-) 0 (-) (-) (-) (-) (-) p.V40M, p.T153N26 (-) 1 (-) (-) (-) (-) (-) (-)

Mutation

Case

Nuclear

β-catenin

accumulatio

n

β-catenin

labeling

index (%)

Table 4. Relationship between GNAS mutation and clinicopathologic factors in GAFG.

Age (years old) 65.60 (60-71) 66.38 (51-79) 0.8524Sex

Male 5 17 0.5552Female 0 4

Locacion Upper third 5 18 0.9999 Middle third 0 3 Lower third 0 0Size of tumor (mm) 26.8 (7-85) 12.571 (3-39) 0.08Macroscopic type

superficial elevated 3 12 0.9999superficial depressed 0 5

Mixed 2 4Depth of submucosal

involvement (μm)720.0 (100-1500) 345.2 (50-1200) 0.0942

Lymphatic invasion 1 (20.0%) 2 (4.7%) 0.4885Venous invasion 0 (0%) 2 (9.5%) 0.9999

Clinicopathologic factorsGNAS mutasion (+)

n=5GNAS mutasion (-)　n=21 p