Portal de Periódicos da CAPES

Molecular Changes Leading to Gastric Cancer: A Suggestion From Rare-Type Gastric Tumors With GNAS Mutations

2014; Elsevier BV; Volume: 146; Issue: 5 Linguagem: Inglês

10.1053/j.gastro.2014.03.025

1528-0012

Kozo Ikuta, Hiroshi Seno, Tsutomu Chiba,

Gastric Cancer Management and Outcomes

Kushima R, Sekine S, Matsubara A, et al. Gastric adenocarcinoma of the fundic gland type shares common genetic and phenotypic features with pyloric gland adenomas. Pathol Int 2013;63:318–325. Recent improvements in hygiene status and eradication of Helicobacter pylori may reduce the incidence of gastric cancer in the near future. However, gastric cancer that develops irrespective of H pylori infection is becoming an emerging issue. Gastric adenocarcinoma of the fundic gland-type (GAFG) is a recently proposed subtype of gastric adenocarcinoma (Am J Surg Pathol 2010;34:609). Most GAFGs develop in nonatrophic gastric mucosa in the absence of H pylori infection, and are composed of irregularly branching and anastomosing tubules lined with basophilic columnar cells with mild nuclear atypia resembling chief cells. Cells positive for H+/K+-ATPase differentiating towards parietal cells are also observed in some GAFGs. On the other hand, pyloric gland adenoma (PGA) is another rare type of gastric tumor that has been increasingly recognized as a distinct variant of gastric-type adenoma (Virchows Arch 2003;442:317). PGA is composed of closely packed pyloric-type glands lined with cuboidal/columnar cells with round nuclei and eosinophilic cytoplasm. Recently, Kushima et al reported that GNAS and KRAS mutations occurring at high frequencies are characteristic genetic features of PGAs of the stomach and duodenum (J Pathol 2013;229:579). GAFGs and PGAs are currently believed to be entirely distinct types of tumors. However, it is becoming increasingly clear that GAFG and PGA share some common clinicopathologic features: Both lesions typically arise in the fundic gland mucosa, occur in older patients, and consistently express MUC6. Based on these findings, Kushima et al extended their previous work by examining the potential relationships between GAFG and PGA, using 3 GAFGs and 12 PGAs (2013;63:318–325). Kushima et al performed immunohistochemical analyses using antibodies for pepsinogen-I, MIST1, H+/K+-ATPase, MUC6, and TFF2 to examine the differentiation statuses of GAFGs and PGAs. In the normal fundic gland mucosa, parietal cells with H+/K+-ATPase are predominant in the upper parts of the glands. Mucous neck cells express pepsinogen-I, MUC6, and TFF2 in the basal regions of the glands, whereas chief cells express pepsinogen-I and the MIST1 transcriptional factor. On the other hand, pyloric gland cells are positive for MUC6 and TFF2 but negative for H+/K+-ATPase, pepsinogen-I, and MIST1. Patients with GAFG were in the age range of 56-78 years (1 male, 2 females; median age, 70 years), and the tumor size was in the range of 4–8 mm (median size, 4). GAFGs were treated by endoscopic submucosal dissection and histologically identified in a background of nonatrophic fundic gland mucosa. No recurrence was detected in any of the patients with GAFG over a follow-up period of 6–63 months. Immunohistochemistry revealed that all 3 GAFG lesions were diffusely positive for pepsinogen-I, MIST1, and MUC6. A small number of H+/K+-ATPase–positive parietal cells were also scattered. TFF2-positive cells were rarely observed in 2 cases, and focally observed in one case. These findings indicate the predominantly mucous neck/chief cell differentiation of these tumors. Patients with PGA were in the age range of 45–89 years (7 males, 5 females; median age, 70 years), and the tumors ranged in size from 5 to 70 mm (median size, 28). All PGAs were identified in a background of fundic gland mucosa, 10 of which were atrophic and were associated with chronic active inflammation and focal pyloric and intestinal metaplasia. The PGAs were diffusely positive for MUC6 and TFF2, suggesting a pyloric gland phenotype. However, focal MIST1 expression was also observed in 10 of the 12 PGAs, and small areas of pepsinogen-I–positive cells were observed in 8 of the 12 PGAs. This observation suggests that PGAs phenotypically resemble mucous neck cells of the fundic gland and show focal chief cell differentiation. For the mutation analyses, Kushima et al microdissected tumor and nontumor areas separately. They subjected extracted DNA to sequencing of exons 8 and 9 of the GNAS gene and exon 2 of the KRAS gene. The mutational analyses identified somatic GNAS mutations in 2 of the 3 GAFGs examined; both were missense mutations, involving codon 201. In contrast with PGAs, no KRAS mutations were observed in any of the GAFGs. Thus, these observations showed the partially overlapping of immunohistochemical profiles and the sharing of GNAS mutations between GAFGs and PGAs. Based on these observations, Kushima et al suggested that both GAFGs and PGAs are closely related lesions characterized by a mucous neck cell/chief cell lineage phenotype in the fundic gland mucosa. GAFG and PGA are considered as entirely distinct tumors. GAFG is defined as an adenocarcinoma, predominantly exhibiting chief cell differentiation. Consistently, Kushima et al observed diffuse expression of 2 chief cell markers, pepsinogen-I and MIST1, in all 3 examined GAFGs. However, it is worth noting that diffuse MUC6 expression and focal TFF2 staining were also observed in GAFGs. On the other hand, all PGAs showed diffuse expression of MUC6 and TFF2, suggesting their pyloric gland phenotype. However, focal expression of pepsinogen-I and MIST1 were also detected in 8 of the 12 PGAs. These findings suggest that GAFG and PGA are closely related lesions characterized by a mucous neck cell/chief cell lineage phenotype in the fundic gland area. Normal fundic gland consists of several different cell types, including mucous neck, parietal, chief, and endocrine cells. It has generally been accepted that these cell lineages differentiate from stem or progenitor cells located in the neck regions of the fundic gland. Stem or progenitor cells in the neck region differentiate and migrate bidirectionally; some move to the luminal side and some toward the basal side of the glands. Cells moving to the base differentiate into mucous neck cells positive for MUC6, TFF2, and pepsinogen-I. During their differentiation into chief cells at the bottom of the glands, mucous neck cells lose their MUC6 and TFF2 expression and obtain MIST1 expression. However, this well-accepted dogma has been challenged by recent fine dissection of the mucous neck/chief cell lineage and also by the recognition of spasmolytic polypeptide-expressing metaplasia (SPEM), a TFF2-expressing metaplastic mucous cell lineage with the phenotypic character of antral gland cells, in Helicobacter-infected gastric mucosa of both mice and humans. Recent genetically engineered mouse studies have revealed the underlying importance of SPEM in the pathologic conditions of the fundic gland. Lineage tracing data using the chief cell marker Mist1-CreERT2 mouse showed that SPEM originates from trans-differentiation of mature chief cells following loss of parietal cells or Helicobacter infection in the stomach (Gastroenterology 2010;139:2028–2037). Differentiated Troy-positive chief cells act as reserve stem cells to generate all lineages of the fundic epithelium, including mucous neck cells (Cell 2013;155:357–368). In these mouse studies, the significance of stem or progenitor cells in the neck regions itself is not necessarily challenged, but under specific conditions chief cells display plasticity to produce all other cell lineages or SPEM through de-differentiation to mucous neck cells. In Kushima et al's study, expression of both MUC6 and chief cell markers in GAFG suggested an intermediate phenotype between mucous neck and chief cells. Furthermore, PGA, which focally expresses pepsinogen-I and MIST1, may also be designated as a tumor with a predominant mucous neck cell phenotype with slight chief cell characteristics rather than a pyloric gland-type tumor. Because the common feature between GAFG and PGA is their mucous neck/chief cell phenotype resembling SPEM rather than simple antralization in the fundic gland area, it may be interesting to speculate that both GAFG and PGA are the tumors, consisting of mucous neck/chief cell expansion, and thus, are on the same differentiation pathway in fundic glands. In this regard, whether the mucous neck/chief cell lineages and/or SPEM are involved in the development of H pylori–associated gastric cancer may also be an important question. In mice, loss of parietal cells caused by Helicobacter infection, drug treatment, or genetic ablation results in gastric mucosa atrophy and SPEM development (Gastroenterology 2010;139:2028–2037). Similarly, atrophy of the corpus and body is associated with the development of SPEM in H pylori-infected humans (reviewed in Gastroenterology 2010;138:2207–2210). Taken together, mucous neck/chief cell expansion and/or SPEM may be important for the pathophysiology of tumors in the fundic gland area, including cancer. An important finding in Kushima's study was that similar to PGA, 2 of the 3 examined GAFGs had mutations of GNAS, a gene encoding the guanine nucleotide-binding protein alpha subunit (Gsα), which is frequently mutated in intraductal papillary mucinous neoplasm of the pancreas and villous adenoma of the colon (Sci Transl Med 2011;3:92ra66; J Pathol 2012;228:113–118). Missense mutations detected in GAFG have been reported as activating mutations (Nature 1989;340:692–696; Pituitary 2007;10:275–282). These results suggest that GAFG and PGA share some similarities in their pathophysiology, although the significance of the activated cyclic AMP-signaling pathway by mutated Gsα remains unknown. In contrast, Kushima et al observed that none of the gastric cancers examined had GNAS mutations (Pathology 2008;40:95–97; J Pathol 2013;229:579–587). These results seem to indicate distinct pathophysiology of GAFG and PGA from H pylori-associated gastric cancer. Furthermore, it should be noted that GAFG typically develops in H pylori-negative, nonatrophic mucosa, whereas PGA is accompanied by atrophic mucosa with chronic inflammation owing to H pylori infection. None of the GAFGs showed KRAS mutations, another common genetic aberration in PGAs (J Pathol 2013;229:579–587). Thus, the relationship among GAFG, PGA, and H pylori-associated gastric cancer with special reference to their ontogeny is an interesting area to be investigated, and we expect that genetically engineered mice experiments, including lineage tracing, will answer this question.