The clinical characteristics of nine patients included in the study are summarized in Table 2. Briefly, the cohort included eight male and one female patient, with a mean age of 72.1 years (range, 58–80 years). Among the patients tested for serum H. pylori IgG antibody (n = 6), five (83%) were positive for the antibody. In one patient who was not tested for the serum antibody, H. pylori was detected in gastric specimen pathological examination. Taken together, six patients were confirmed to be infected by H. pylori. None of the patients had a history of long-term proton pump inhibitor (PPI) use or clinical features suggesting of autoimmune gastritis. Serum gastrin levels were not measured. One tumor (case 5) was from a patient with FAP, whereas the remaining tumors were not associated with a genetic predisposition such as FAP, multiple endocrine neoplasia (MEN), or Lynch syndrome.
Table 2 Clinical information of the patients included in the studyThe tumors were mostly (6/9) located in the body of the stomach. Endoscopically, all nine tumors appeared as flat elevated lesions. The tumor size ranged from 0.8 to 4.4 cm.
Endoscopic submucosal dissection was performed in seven patients, and one tumor was surgically resected. In one patient (case 9), only biopsy specimens were available, because the patient died due to another disease before gastric tumor resection.
In cases 4 and 7, the time interval between diagnosis and resection was long (60 and 52 months, respectively). During this period, there was a minimal change in tumor size between the first examination and resection in case 7, whereas the tumor increased from 3 to 4 cm in case 4.
Pathologic featuresThe histologic features of all nine cases are summarized in Table 3. In all cases, tumors were intramucosal lesions with a superficially localized adenoma component and a more deeply localized NET component. The adenomatous component was composed of columnar epithelium with elongated nucleus and pseudostratification in all nine cases. The adenomatous component was uniformly low-grade in six cases but was adenoma exhibiting predominantly low-grade and focal high-grade dysplasia in the remaining three cases.
Table 3 Pathologic findings of the cases in the present studyThe NET component was composed of uniform cells arranged in small nests or cords. NET components were also noted within the adenomatous glands at the basal side or were budded from the adenomatous glands. Neuroendocrine cells had nuclei with a salt-and-pepper pattern and indistinct nucleoli. The NET component did not exhibit mitosis or necrosis in any of the cases. Two intratumoral distribution patterns of the NET component were noted: (i) distribution from middle to deep lamina propria (basal–middle pattern; Fig. 1; n = 6) and (ii) distribution confined to deep lamina propria under the adenomatous component (deep lamina propria pattern; Fig. 2; n = 3). In two cases (cases 4 and 7), the horizontal spreading of the NET component was broader than that of the adenomatous component. The distribution pattern (basal–middle vs. deep lamina propria) was not associated with horizontal extension of the NET component or with tumor diameter.
Fig. 1Representative images of gastric MANET, basal–middle pattern. A, B The neuroendocrine tumor (NET) component is seen in the middle/deep layer of lamina propria (basal–middle pattern) (A). Synaptophysin staining (B) highlights the NET component. C Higher magnification of the adenomatous component. D Higher magnification of the NET component forming small nests. E Note that clusters of neuroendocrine cells (arrows) are protruding from adenomatous glands. (A–D case 2; E case 1). MANET, mixed adenoma-neuroendocrine tumor
Fig. 2Representative images of gastric MANET, deep lamina propria pattern. A The NET component is confined to deep lamina propria (deep lamina propria pattern). B Chromogranin A staining highlights the NET component in deep lamina propria. C Higher magnification of the adenoma component (low-grade tubular adenoma) composed of closely packed neoplastic glands with elongated nuclei. D Higher magnification of the NET component. Small clusters of neuroendocrine cells are seen immediately above muscularis mucosa. They are positive for serotonin (D, inset). (All sections from case 5)
None of the eight resected tumors exhibited vascular invasion. In all cases, the nonneoplastic background mucosa was gastric body mucosa exhibiting chronic atrophic gastritis with moderate/marked intestinal metaplasia and mucosal atrophy, a status compatible with the consequence of H. pylori gastritis. None of the cases had histopathological findings suggestive of autoimmune gastritis, such as enterochromaffin-like cell (ECL cell) hyperplasia. PPI-related changes, such as parietal cell protrusion or oxyntic gland dilataion, were also absent in the background mucosa.
Immunohistochemical findingsImmunohistochemical analyses revealed that the NET component was diffusely positive for both chromogranin A and synaptophysin in all nine cases. In addition, these immunostainings highlighted linear or scattered neuroendocrine cells at the basal side of the adenoma component in eight of the nine cases. In six of the nine cases, the NET component was budding from the adenomatous component (Fig. 1E). In six of the seven assessed tumors, the NET component was diffusely positive for serotonin (Fig. 2D, inset) and showed an expression pattern of SSTR2A + /CDX2 + /VMAT2 − (Supplementary Fig. 1), suggesting they are serotonin-producing enterochromaffin-cell (EC-cell) phenotype [2] (Table 3, right). The NET component of a serotonin-negative case also showed a pattern of SSTR2A + /CDX2 + /VMAT2 − , which suggests EC-cell phenotype, though it lacked positivity for all the assessed hormones (serotonin, gastrin, somatostatin).
In all cases, the Ki-67 proliferation index was less than 2% in the NET component and was ranged between 2 and 30% in the adenomatous component. Regarding the cellular phenotype of the adenomatous component, six cases were classified as the intestinal type and two cases were classified as the mixed type (predominantly intestinal phenotype).
Wild-type p53 staining pattern was observed in all but one case (case 8), in which mutant p53 pattern with diffuse and strong staining was noted in the high-grade adenoma component (Fig. 3). In none of the cases, nuclear β-catenin expression was observed in the adenomatous or the NET components.
Fig. 3A case of gastric MANET with focal pattern of mutant p53 staining—case 8. A–C Loupe view of a section containing low-grade (left to center) and focal high-grade dysplasia component (right). Hematoxylin/eosin staining (A) shows that the area with high-grade dysplasia (right) is slightly depressed compared to the area with low-grade dysplasia. Chromogranin A staining (B) highlights the distribution of the NET component. C Immunohistochemistry for p53. Most adenoma glands (left) show wild-type p53 pattern, whereas focal high-grade dysplasia component (right) and adjacent periphery of low-grade component (center) show diffuse p53 positivity. The NET component shows wild-type p53 staining pattern. D Higher magnification of the MANET. Small vesicles composed of neuroendocrine cells intermingle with adenomatous glands in deep lamina propria. E Higher magnification of the focal high-grade dysplasia component composed of irregular glands with enlarged ovoid nuclei
Review of the biopsy specimensPreoperative biopsy specimens were available for review in six patients who underwent resection. Both the NET and adenomatous components were recognized in the biopsy specimens of two of the six cases (Fig. 4); in both cases, the NET component exhibited the basal–middle pattern of distribution.
Fig. 4Prior gastric biopsy specimen of a patient later diagnosed with gastric MANET, basal–middle pattern—case 7. A Whole section of the biopsied specimen. The NET component, which forms nests smaller than adenomatous glands, is located in the center. B Higher magnification shows the NET component (inset, synaptophysin) together with adenoma glands
Follow-up dataFollow-up information was available in all nine patients, with follow-up duration ranging from 2 months to 8 years (mean, 30 months). One patient (case 9) died due to advanced colon cancer with urinary bladder invasion 2 months after the gastric biopsy. All of the remaining eight patients were alive at last-follow-up with no recurrence (Table 4).
Table 4 Summary of the literature review, including the patients with gastric MANET included in this study
Comments (0)