Indexed In : Science Citation Index Expanded(SCIE), MEDLINE,
Pubmed/Pubmed Central, Elsevier Bibliographic, Google Scholar,
Databases(Scopus & Embase), KCI, KoreaMed, DOAJ
Gut and Liver is an international journal of gastroenterology, focusing on the gastrointestinal tract, liver, biliary tree, pancreas, motility, and neurogastroenterology. Gut atnd Liver delivers up-to-date, authoritative papers on both clinical and research-based topics in gastroenterology. The Journal publishes original articles, case reports, brief communications, letters to the editor and invited review articles in the field of gastroenterology. The Journal is operated by internationally renowned editorial boards and designed to provide a global opportunity to promote academic developments in the field of gastroenterology and hepatology. +MORE
Yong Chan Lee |
Professor of Medicine Director, Gastrointestinal Research Laboratory Veterans Affairs Medical Center, Univ. California San Francisco San Francisco, USA |
Jong Pil Im | Seoul National University College of Medicine, Seoul, Korea |
Robert S. Bresalier | University of Texas M. D. Anderson Cancer Center, Houston, USA |
Steven H. Itzkowitz | Mount Sinai Medical Center, NY, USA |
All papers submitted to Gut and Liver are reviewed by the editorial team before being sent out for an external peer review to rule out papers that have low priority, insufficient originality, scientific flaws, or the absence of a message of importance to the readers of the Journal. A decision about these papers will usually be made within two or three weeks.
The remaining articles are usually sent to two reviewers. It would be very helpful if you could suggest a selection of reviewers and include their contact details. We may not always use the reviewers you recommend, but suggesting reviewers will make our reviewer database much richer; in the end, everyone will benefit. We reserve the right to return manuscripts in which no reviewers are suggested.
The final responsibility for the decision to accept or reject lies with the editors. In many cases, papers may be rejected despite favorable reviews because of editorial policy or a lack of space. The editor retains the right to determine publication priorities, the style of the paper, and to request, if necessary, that the material submitted be shortened for publication.
Yuri Kim1 , Bokyung Ahn2 , Kee Don Choi1 , Beom-Su Kim3 , Jeong-Hwan Yook3 , Gin Hyug Lee1 , Seung-Mo Hong2 , Jeong Hoon Lee1
Correspondence to: Seung-Mo Hong
ORCID https://orcid.org/0000-0002-8888-6007
E-mail smhong28@gmail.com
Jeong Hoon Lee
ORCID https://orcid.org/0000-0002-0778-7585
E-mail jhlee.gi@amc.seoul.kr
Yuri Kim and Bokyung Ahn are contributed equally to this work as first authors.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Gut Liver 2023;17(6):863-873. https://doi.org/10.5009/gnl220175
Published online January 2, 2023, Published date November 15, 2023
Copyright © Gut and Liver.
Background/Aims: Although gastric neuroendocrine tumors (NETs) are uncommon neoplasms, their prevalence is increasing. The clinical importance of the World Health Organization (WHO) classification of gastric NETs, compared with NETs in other organs, has been underestimated. This study aimed to systematically evaluate the clinical and pathologic characteristics of gastric NETs based on the 2019 WHO classification and to assess the survival outcomes of patients from a single-center with a long-term follow-up.
Methods: The medical records of 427 patients with gastric NETs who underwent endoscopic or surgical resection between January 2000 and March 2020 were retrospectively reviewed. All specimens were reclassified according to the 2019 WHO classification. The clinicopathologic characteristics, treatment, and oncologic outcomes of 139 gastric NETs were analyzed.
Results: The patients’ median age was 53.0 years (interquartile range [IQR], 46.0 to 63.0 years). The median follow-up period was 36.0 months (IQR, 15.0 to 63.0 months). Of the patients, 92, 44, and 3 had grades 1, 2, and 3 NETs, respectively. The mean tumor size significantly increased as the tumor grade increased (p=0.025). Patients with grades 2 and 3 gastric NETs more frequently had lymphovascular invasion (29.8% vs 10.9%, p=0.005) and deeper tissue invasion (8.5% vs 0%, p=0.012) than those with grade 1 tumors. The overall disease-specific survival rate was 100%. Two patients with grades 2-3 gastric NETs experienced extragastric recurrence.
Conclusions: Although gastric NETs have an excellent prognosis, grade 2 or grade 3 gastric NETs are associated with a larger size, deeper invasion, and extragastric recurrence, which require active treatment.
Keywords: Neuroendocrine tumor, Stomach, World Health Organization, Classification, Survival
Since Oberndorfer1 first described the term “karzinoide” in 1907, neuroendocrine tumors (NETs) have been consistently found in the different organs of the gastrointestinal (GI) tract, and their prevalence has been gradually increasing owing to the increasing number of patients who undergo routine medical examinations.1 Gastric NETs are rare types of tumors, with an overall incidence of 0.2 cases per 100,000 people and a prevalence of 0.6% to 2% of all GI NETs.2-5
The most common gastric NETs are enterochromaffin-like (ECL)-cell (histamine-producing) tumors.6,7 In 1993, Rindi et al.8 first classified gastric ECL-cell NETs into three subtypes based on the histology of the adjacent mucosa, presence or absence of antral G-cell hyperplasia and hypergastrinemia, and accompanying clinical conditions. Type I gastric ECL NETs are the most common subtype. These are multiple small tumors occurring in the body or fundus and are associated with hypergastrinemia, G-cell hyperplasia, and autoimmune chronic atrophic gastritis. Type II ECL NETs develop in the body and fundus, with hypertrophy of the adjacent mucosa, and measure <2 cm in size. These tumors occur in patients with multiple endocrine neoplasia type 1 or Zollinger-Ellison syndrome and are accompanied by multiple endocrine neoplasms in the parathyroid gland, pituitary gland, and pancreas.9 Lastly, type III ECL NETs are sporadic tumors measuring >2 cm in size. They present as a single mass in any part of the stomach and are not associated with hypergastrinemia, autoimmune chronic atrophic gastritis, multiple endocrine neoplasia type 1, or Zollinger-Ellison syndrome.7,9,10 This classification method is only used for gastric NETs and not for NETs in other organs, with no substantial change since it was introduced in 1993.
Unlike the tumors in other organs, stomach tumors were distinguished using not only the World Health Organization (WHO) classification system but also the classification methods according to types. The 2019 WHO classification of gastric NETs was updated recently to better stratify the patients’ prognosis according to the following grading system: grade 1 (G1), <2 mitoses/2 mm2 in 10 high-power fields (HPFs) at 40× magnification and an ocular field diameter of 0.5 mm, and Ki-67 labeling index (Ki67LI) of <3%; grade 2 (G2), 2–20 mitoses/2 mm2 or Ki67LI of 3% to 20%; and grade 3 (G3), ≥20 mitoses/2 mm2 or Ki67LI of >20%.11 Although a few previous studies investigating gastric NETs have been published, they only included a small number of cases, lack uniformity in classifying the cases using the previous WHO criteria, and lack clinicopathologic or survival data.12-14 Therefore, this study aimed to systematically evaluate the clinical characteristics and pathologic findings of gastric NETs subclassified according to the updated 2019 WHO classification in a large-scale cohort of patients.
This single-center, retrospective study was conducted in Asan Medical Center, Seoul, Korea. A total of 427 patients diagnosed with NETs, well-differentiated NETs, poorly differentiated NETs, well-differentiated neuroendocrine neoplasms, well-differentiated neuroendocrine carcinoma (NEC), poorly differentiated NEC, and carcinoid tumors between January 2000 and March 2020 were recruited for this study. Fig. 1 shows a flowchart demonstrating the inclusion and exclusion criteria used to screen the initial cohort. A total of 139 patients with well-differentiated gastric NETs were finally selected for subsequent analysis.
An endoscopic evaluation was performed to assess the characteristics of NETs (including the size, number, and location of tumors) and the surrounding gastric mucosa. The location of gastric NETs was divided into the upper, middle, lower, and multiregional parts of the stomach. The upper part included the fundus, cardia, and high body; the middle part included the mid and lower body; and the lower part included the gastric angle and antral portion. Atrophy of the gastric mucosa was evaluated and classified as either closed or open type according to the Kimura-Takemoto classification.15 Intragastric recurrence was classified as synchronous recurrence and metachronous recurrence. Synchronous gastric NETs were defined as new gastric NETs additionally diagnosed within 6 months after the previous diagnosis, while metachronous gastric NETs were defined as new lesions detected after 6 months had elapsed.16 Extragastric recurrence was defined as tumor recurrence in organs other than the stomach after endoscopic or surgical resection of gastric NETs.
All patients underwent endoscopic (n=116) or surgical (n=23) resection. Although there are no clear therapeutic criteria, endoscopic resection was considered first when the size of the lesion was ≤2 cm, with no evidence of extragastric metastasis, and when the lesion was localized in the mucosal or submucosal layer on endoscopic ultrasonography. In contrast, when the size of the lesion was relatively large, or was accompanied by a deep ulcer on the surface making endoscopic resection difficult, or was deeper than the submucosal layer in endoscopic ultrasonography, or when the patient desired surgery, surgical treatment was performed. The different endoscopic treatments included endoscopic submucosal dissection (ESD), endoscopic mucosal resection, and argon plasma coagulation. The endoscopic procedures were performed by six expert endoscopists. Each endoscopic procedure was carefully selected and performed at the discretion of the endoscopists and according to the characteristics of the gastric NETs. The surgical methods performed were wedge resection, distal gastrectomy, and total gastrectomy. With regard to the endoscopic treatments, the surgical methods and extent of resection were determined by the GI surgeons. Follow-up esophagogastroduodenoscopy (EGD) or abdominal computed tomography examinations were performed every 6 to 12 months for the first 2 years after the endoscopic or surgical treatment and annually thereafter.
All hematoxylin and eosin-stained slides were carefully re-evaluated by two pathologists who were blinded to the patients’ clinicopathologic information. The tumor grade was evaluated using mitotic counts/10 HPFs and Ki67LI according to 2019 WHO classification system.11 Briefly, G1 was assigned to tumors with <2 mitoses/10 HPFs or Ki67LI of <3%; G2, to tumors with 2–20 mitoses/10 HPFs or Ki67LI of 3% to 20%; and G3, to tumors with >20 mitoses/10 HPFs or Ki67LI of >20%.11 Each tumor slide was manually scanned under a ×10 objective, and the area with the greatest Ki-67 positivity (“hotspot”) was selected for image capturing and printing. If the hotspot included <500 cells (the recommended minimum count of tumor cells for accurate grading), ≥2 HPFs were additionally selected, if available, to include a sufficient number of tumor cells.11 MicroNETs were defined as nodules of ECL cells with a diameter ranging from 0.05 to 0.5 cm.17 Neuroendocrine dysplasia (micronodules of neuroendocrine cells measuring >150 µm and <0.5 mm in size), NECs (small cell carcinomas or large cell carcinomas with sheets or poorly formed trabeculae or nests of neuroendocrine cells), and mixed neuroendocrine and non-neuroendocrine neoplasms (neoplasms with morphologically recognizable neuroendocrine and non-neuroendocrine components, each constituting ≥30% of the tumor volume) were excluded before and after central pathology review.12,18 Lymphovascular and perineural invasions, marginal status, and lymph node and distant metastatic status were also evaluated. The tumor stage was evaluated based on the 8th edition of the American Joint Committee on Cancer Staging Manual.18
Thirty-one cases with no available Ki67LI data were selected for Ki-67 immunohistochemistry. Immunohistochemical staining of the tumor tissue samples expressing Ki-67 was performed using the BenchMark XT autoimmunostainer (Ventana Medical Systems, Tucson, AZ, USA) with the OptiView DAB Detection Kit (Ventana Medical Systems) according to the manufacturer’s instructions and using the reagents supplied along with the kit. Briefly, 4-µm sections were mounted on silanized charged slides, allowed to dry for 10 minutes at room temperature, and placed in an incubator for 20 minutes at 65°C. The sections were treated with heat induced epitope retrieval (CC1) for 64 minutes and incubated for 32 minutes with mouse anti-Ki-67 antibody (clone MIB1, dilution 1:200; DAKO, Glostrup, Denmark) in the autoimmunostainer.
The baseline characteristics of gastric NETs subclassified according to the tumor grade were analyzed. Quantitative data (e.g., patients’ age, follow-up duration, and tumor size) were expressed as medians with the interquartile range. Qualitative data (e.g., patients’ sex; symptoms; multiorgan involvement; and number, location, atrophy, and pathologic characteristics of tumor specimens) were expressed as proportions. The categorical variables were compared using the non-parametric Kruskal-Wallis test. Owing to the small number of G3 gastric NETs, G2 and G3 NETs were grouped together (G2-3) and were compared with G1 gastric NETs. The clinical characteristics and outcomes of patients with microNETs were separately analyzed by conducting a subgroup analysis. A probability (p) value of <0.05 was considered significant. All statistical analyses were performed using IBM SPSS Statistics for Windows version 24.0 (IBM Corp., Armonk, NY, USA) and Prism 9 for Windows version 9.3.1 (GraphPad Software Inc., San Diego, CA, USA).
The study was approved by the Institutional Review Board of Asan Medical Center (IRB number: 2020-0996). The requirement for informed consent was waived because the data were retrospectively collected. All procedures were performed in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the 1964 Declaration of Helsinki and its later versions.
The grades of all specimens were re-evaluated according to the updated 2019 WHO classification system through a centralized pathology review performed by expert pathologists; this resulted in changes in gastric NET grade in some cases. After the central pathology review, 139 patients were re-evaluated. The representative endoscopic and pathologic images of gastric NETs by tumor grade are depicted in Fig. 2.
The baseline characteristics of the patients are summarized in Table 1. Of the 139 patients, 92 (66.2%), 44 (31.7%), and three (2.1%) had G1, G2, and G3 NETs, respectively. Most patients were asymptomatic (62.0% of those with G1 NETs and 70.5% of those with G2 NETs), and gastric NETs were often incidentally detected during routine EGD examinations. However, all three patients with G3 NETs experienced abdominal pain prior to the diagnosis.
Table 1. Patients’ Baseline Characteristics and Endoscopic Findings
Characteristic | Total (n=139) | Grade 1 (n=92) | Grade 2 (n=44) | Grade 3 (n=3) | p-value* |
---|---|---|---|---|---|
Age, yr | 53.0 (46.0–63.0) | 53.5 (46.0–63.0) | 51.5 (43.3–60.0) | 59.0 (52.0–59.0) | 0.311 |
Follow-up period, mo | 36.0 (15.0–63.0) | 31.5 (13.5–61.8) | 42.5 (22.8–68.0) | 35.0 (15.0–35.0) | 0.481 |
Sex | 0.765 | ||||
Male | 74 (53.2) | 50 (54.3) | 23 (52.3) | 1 (33.3) | |
Female | 65 (46.8) | 42 (45.7) | 21 (47.7) | 2 (66.7) | |
Symptom | 0.077 | ||||
None | 88 (63.3) | 57 (62.0) | 31 (70.5) | 0 | |
Pain | 46 (33.1) | 33 (35.8) | 10 (22.7) | 3 (100.0) | |
Bleeding | 2 (1.4) | 0 | 2 (4.5) | 0 | |
Carcinoid symptom | 3 (2.2) | 2 (2.2) | 1 (2.3) | 0 | |
Multiorgan involvement | 6 (4.2) | 4 (4.3) | 2 (4.6) | 3 (100.0) | 0.931 |
Tumor size, mm | 10.0 (6.0–10.0) | 8.0 (5.0–10.0) | 10.0 (6.0–15.0) | 15.0 (10.0–15.0) | 0.025† |
Number | 0.558 | ||||
Single | 109 (78.4) | 73 (79.3) | 33 (75.0) | 3 (100.0) | |
Multifocal | 30 (21.6) | 19 (20.7) | 11 (25.0) | 0 | |
Location | 0.042† | ||||
Upper | 47 (33.8) | 32 (34.8) | 15 (34.1) | 0 | |
Middle | 71 (51.1) | 49 (53.2) | 22 (50.0) | 0 | |
Lower | 7 (5.0) | 3 (3.3) | 1 (2.3) | 3 (100.0) | |
Multiregional | 14 (10.1) | 8 (8.7) | 6 (13.6) | 0 | |
Atrophy | 0.588 | ||||
None | 23 (16.5) | 17 (18.5) | 5 (11.4) | 1 (33.3) | |
Closed type | 19 (13.7) | 13 (14.1) | 6 (13.6) | 0 | |
Open type | 97 (69.8) | 62 (67.4) | 33 (75.0) | 2 (66.7) | |
Peptic ulcer disease | 7 (5.0) | 3 (3.3) | 2 (4.5) | 2 (66.7) | <0.001† |
Data are presented as median (interquartile range) or number (%).
*Statistical significance was assessed using the Kruskal-Wallis test; †p<0.05.
The endoscopic findings of each tumor grade are summarized in Table 1. The median macroscopic size of gastric NETs was significantly larger in patients with higher-grade tumors (G1 vs G2 vs G3, 8.0 mm vs 10.0 mm vs 15.0 mm, p=0.025). A total of 71 tumors (51.1%) were detected in the middle part of the stomach, with all G3 NETs located in the lower part (53.3%, 50.0%, and 0% for G1, G2, and G3, respectively, p=0.042). Peptic ulcers rarely occurred in patients with G1 and G2 NETs; however, two of the three G3 NETs were detected in the form of multiple ulcers (3.3% vs 4.5% vs 66.7%, p<0.001).
Owing to the small number of patients with G3 tumors, those with G2 and G3 tumors were merged and analyzed as one group. A total of 119 patients (85.6%) underwent at least one endoscopic procedure (Table 2, Supplementary Fig. 1). Eleven patients received repeated endoscopic procedures. Patients with a synchronous lesion size with a size or shape change (eight patients) or a metachronous lesion (one patient) and suspected recurrence at the previous resected site (one patient) underwent repeated endoscopic treatment. The number of patients who underwent surgical treatment was significantly higher in the G2-3 NET group than in the G1 NET group (10.9% for G1 vs 38.3% for G2-3, p<0.001). Lymph node metastasis was confirmed in six of 22 patients who underwent surgical treatment (one patient with G1 NET and five with G2 NETs). Lymphovascular invasion was more frequently observed in patients with G2-3 NETs than in those with G1 NETs (10.9% for G1 vs 29.8% for G2-3, p=0.005). However, no significant differences were observed in perineural invasion and resection margin status. Most of the patients had gastric NETs in the mucosal or submucosal layer, whereas serosal invasion was confirmed by pathologic review in four patients with G2 NETs (2.9%). Additional chemotherapy treatment was more frequently administered in patients with G2-3 NETs than in those with G1 NETs (1.1% for G1 vs 10.7% for G2-3, p=0.009). When the treatment results were analyzed according to the resection method, the specimens obtained through surgical resection had a larger size, higher pathological grade and negative resection margin rates, lower rates for lymphovascular invasion and perineural invasion, and deeper invasion depth (Supplementary Table 1).
Table 2. Comparison of Treatment and Oncologic Outcomes
Variable | Total (n=139) | Grade 1 (n=92) | Grade 2-3 (n=47) | p-value* |
---|---|---|---|---|
Endoscopic treatment | 0.058 | |||
None | 20 (14.4) | 8 (8.7) | 12 (25.5) | |
1 Time | 109 (78.4) | 78 (84.8) | 31 (66.0) | |
2 Times | 8 (5.8) | 4 (4.3) | 4 (8.5) | |
3 Times | 2 (1.4) | 2 (2.2) | 0 | |
Surgery | <0.001† | |||
None | 111 (79.9) | 82 (89.1) | 29 (61.7) | |
Wedge resection | 11 (7.9) | 5 (5.4) | 6 (12.8) | |
Subtotal gastrectomy | 5 (3.6) | 1 (1.1) | 4 (8.5) | |
Total gastrectomy | 12 (8.6) | 4 (4.3) | 8 (17.0) | |
Pathologic size, mm | 8.0 (6.0–12.0) | 7.0 (4.3–9.0) | 11.0 (6.0–15.0) | <0.001† |
Lymphovascular invasion | 0.005† | |||
Negative | 115 (82.7) | 82 (89.1) | 33 (70.2) | |
Positive | 24 (17.3) | 10 (10.9) | 14 (29.8) | |
Perineural invasion | 0.162 | |||
Negative | 138 (99.3) | 92 (100.0) | 46 (97.9) | |
Positive | 1 (0.7) | 0 | 1 (2.1) | |
Resection margin | 0.210 | |||
Negative | 117 (84.2) | 80 (87.0) | 37 (78.7) | |
Positive | 22 (15.8) | 12 (13.0) | 10 (21.3) | |
Invasion depth | 0.012† | |||
Mucosa | 30 (21.6) | 24 (26.1) | 6 (12.8) | |
Submucosa | 105 (75.5) | 68 (73.9) | 37 (78.7) | |
Muscularis propria | 0 | 0 | 0 | |
Serosa | 4 (2.9) | 0 | 4 (8.5) | |
Chemotherapy | 0.009† | |||
None | 133 (95.7) | 91 (98.9) | 42 (89.4) | |
Adjuvant | 3 (2.2) | 1 (1.1) | 2 (4.2) | |
Palliative | 3 (2.2) | 0 | 3 (6.4) | |
Intragastric recurrence | 0.966 | |||
None | 107 (77.0) | 71 (77.2) | 36 (76.6) | |
Synchronous | 31 (22.3) | 20 (21.7) | 11 (23.4) | |
Metachronous | 1 (0.7) | 1 (1.1) | 0 | |
Extragastric recurrence | 0.047† | |||
None | 137 (98.6) | 92 (100.0) | 45 (95.7) | |
Recurred | 2 (1.4) | 0 | 2 (4.3) | |
All-cause mortality | 0.754 | |||
Death | 9 (6.5) | 5 (5.4) | 4 (8.5) | |
Survived | 130 (93.5) | 87 (94.6) | 43 (91.5) |
Data are presented as number (%) or median (interquartile range).
*Statistical significance was assessed using the Kruskal-Wallis test; †p<0.05.
In patients who underwent endoscopic procedures, 15 patients were lymphovascular invasion positive. Among them, five patients underwent additional surgical treatment (two wedge resections, two subtotal gastrectomy, and one total gastrectomy). This decision was based on the depth of submucosal invasion, positive resection margin in addition to lymphovascular invasion positivity, or the patient’s desire for additional surgical treatment. There was no lesion recurrence in any of the patients, regardless of whether they underwent additional surgery.
There was one patient who had metachronous gastric NET. A 44-year-old female underwent ESD for a 12-mm-sized submucosal tumor-like NET on the posterior wall of the mid body. Seven years after the first procedure, a 3-mm sized G1 NET was newly detected in the posterior wall of the high body, and additional endoscopic mucosal resection was performed. After two procedures, no additional recurrence was found during regular follow-up endoscopy.
Two patients (1.4%) experienced extragastric recurrence (Table 2). One patient was a 40-year-old woman who underwent total gastrectomy for a 30-mm gastric NET in the mid body, which was histologically confirmed as a G2 tumor. Four months after surgery, peritoneal carcinomatosis was detected on follow-up abdominal computed tomography; hence, palliative irinotecan chemotherapy was administered. The other patient was a 52-year-old woman who experienced epigastric discomfort and had a 10-mm NET in the antrum, which was detected during EGD and histologically confirmed as a G3 tumor with no lymphovascular or perineural invasion and with a clear resection margin. Follow-up EGD performed 32 months after the initial treatment showed G3 NET recurrence at the post-ESD site, with multiple bones and liver metastases identified on abdominal computed tomography images. The patient was treated with lanreotide and everolimus and underwent combination chemotherapy with cisplatin and fluorouracil.
Nine patients (6.5%) died during the study (five patients with G1 NETs and four patients with G2 NETs), with no significant survival difference between the groups (94.6% for G1 vs 91.5% for G2-3, p=0.754). None of the nine deaths were due to the progression of gastric NETs; six patients died of progression of other underlying diseases, while three patients died of unknown causes (Supplementary Table 2).
Of the 21 patients with microNETs (Figs 3-5), most had G1 NETs (16 cases, 76%) (Figs 4 and 5), whereas four patients had G2 NETs despite the small tumor size (Fig. 3). No significant differences were found in the clinical characteristics between the two groups. Five patients (23.8%) experienced intragastric recurrences during the observation period; however, none had extragastric recurrence (Supplementary Table 3).
Our study aimed to systematically re-evaluate the clinicopathologic characteristics and survival outcomes of patients with gastric NETs subclassified according to the updated 2019 WHO classification system. Previous three-tiered classifications of gastric NETs, unlike those of NETs in other organs, were mainly based on the presence or absence of atrophy and hypergastrinemia. However, such classification methods were limited by their ambiguity or the practical difficulties in conducting tests (e.g., gastrin test) in the clinical setting. Therefore, to compare the clinical characteristics and prognosis based on the pathologic classification alone, similar to the classification of NETs in other organs, this study was conducted after excluding type-related factors. In the present study, no significant differences were found in the patients’ age and sex among the tumor grades; however, a higher grade was found to be correlated with a larger tumor, higher rate of lymphovascular invasion, and a greater depth of invasion, even after excluding the factors related to the types of gastric NETs. Considering these results, the 2019 WHO classification alone may be sufficient in predicting the clinical features and prognosis of gastric NETs. Because our cohort consisted of patients diagnosed with gastric NETs within the last 20 years, when various terminologies were used with inaccurate tumor grade, lack of Ki-67 immunohistochemistry or inaccurate Ki67LI owing to “eyeballing” measurements, many of the patients were reclassified after a centralized review. In the process, cases that did not meet the current diagnostic criteria for gastric NETs were excluded from the study, thus increasing the accuracy of our results.
The prevalence of NETs has been gradually increasing owing to an increase in the number of patients undergoing routine medical evaluations.1,19 According to a number of studies the age-adjusted incidence of GI NETs in 1973 was 1.09 per 100,000, which significantly increased to 5.25 per 100,000 in 2004 and 6.98 per 100,000 in 2012 in the United States.20,21 The NETs that commonly occur in the GI tract are small-intestinal NETs (30.8%), followed by rectal (26.3%), colonic (17.6%), and pancreatic (12.1%) NETs. By contrast, gastric NETs are relatively infrequent (8.9%).22,23 Unlike the NETs occurring in other GI organs, which are classified solely based on pathologic grading, gastric NETs have been categorized according to the presence or absence of hypergastrinemia or atrophy.9 However, these classification methods are often difficult to apply in real-world clinical settings, especially in primary clinics, owing to the difficulties in measuring the serum gastrin levels and objectively evaluating gastric atrophy. Because gastrin is released by G cells in response to the stimulation of gastrin-releasing peptides in the stomach, and a person’s food intake or prolonged use of proton pump inhibitors can lead to fluctuations in serum gastrin levels, the results of the serum gastrin test should be interpreted with caution.24-26 Even if the gastrin levels and atrophy are assessed, several patients do not meet the criteria for ECL NET subtyping. Therefore, applying the conventional ECL-cell NET subtyping criteria is often ambiguous. In addition to this subtyping, the WHO classification of NETs based on the Ki-67 index and mitotic status was defined in 2010 and was recently revised in 2019.11,27 Ki-67 index was added to the classification criteria after Garcia-Carbonero et al.28 found that Ki-67 is an independent predictor of survival. However, only a few studies have determined the prognostic significance of classifying gastric NETs according to the 2019 WHO classification system, compared with ECL-cell NET typing; thus, its importance is relatively underestimated.
In this study, all pathologic specimens, including those diagnosed before 2010, were centrally reviewed and analyzed together; most gastric NETs were G1 tumors (66.2%) of <10 mm in size and developed in the upper part of the stomach. These characteristics are consistent with the features of type I ECL NETs, which account for 70% to 80% of all NETs (i.e., relatively small in size and often found in the fundus and body of the stomach).14,29 Although the all-cause mortality rate was reported to be 6.5%, the disease-specific mortality rate was 0%, and the number of patients with recurrent gastric NETs was extremely small. Thus, gastric NETs are considered to have an excellent overall prognosis. However, as the tumor grade increased, the gastric NETs showed a gradual increase in size and a greater association with lymphovascular invasion after endoscopic or surgical resection. Therefore, in patients with G2 or G3 gastric NETs, close observation is necessary owing to their higher risk of developing local recurrence after undergoing endoscopic or surgical treatment compared with those with G1 NETs. Clinical data on the recurrence rate and average recurrence period of gastric NETs were lacking, with no consensus on the follow-up period after treatment. Of the 139 patients, one patient with a G2 gastric NET experienced recurrence 4 months after surgery, while one patient with a G3 gastric NET experienced recurrence at 32 months after ESD, with a median follow-up of 36 months. Considering these findings, these patients should be closely observed for 5 years, which is similar to that of patients with gastric adenocarcinoma. In this regard, guidelines based on the results of further studies should be established.
All patients included in this study underwent surgery or endoscopic treatment, and those who had undergone endoscopic forceps biopsy alone were excluded owing to the extremely small specimen size. Because NETs are indolent and have a lower malignancy risk compared with other GI tumors, the National Comprehensive Cancer Network and European Neuroendocrine Tumor Society guidelines only recommend surveillance if the lesion is confined to the mucosal or submucosal layer or <10 mm in size.30-34 However, our study showed that even in patients with microNETs, unlike those with neuroendocrine microadenomas in the pancreas, higher-grade tumors (e.g., G2) may be left undetected.35,36 Furthermore, some experts suggested that it may be better to remove small gastric NETs using endoscopic methods, such as polypectomy, endoscopic mucosal resection, and ESD.37-40 Considering the results of these studies, if a gastric NET is diagnosed, active treatment using endoscopic or surgical methods should be considered, even for those with small gastric NETs.
This study has several limitations. First, the results may be limited by the retrospective and single-center study design. In addition, as patients with small gastric NETs who only underwent observation, those whose tumors were removed by forceps biopsy, and those without proper endoscopic or surgical specimens were excluded, selection bias may be present. Third, the number of patients with G3 NETs was relatively small, which might have confounded the analysis. However, three patients (2.1%) had G3 gastric NETs; this prevalence rate is consistent with that reported in previous studies, although the data on G3 NETs are limited owing to their recent distinction from NECs in 2017.41-43 In the PRONET study of neuroendocrine neoplasms in the lungs and digestive system, 21 patients (2.7%) had G3 NETs among 778 patients with GI neuroendocrine neoplasms, which was similar to the prevalence of G3 NETs in our study.41-43 Although the PRONET study differs from our study in that it examined all types of NETs, it provides evidence that G3 NETs are rare. Therefore, further studies with a larger number of patients with gastric NETs, including those who underwent observation or endoscopic biopsy alone, are needed to obtain more accurate results, including those of patients with G3 NETs. Finally, as two patients experienced recurrence at 4 and 32 months after the initial treatment of gastric NETs, and the median follow-up period of the overall patients was 36 months, the follow-up period in this study was relatively short. As mentioned above, for accurate observation, a 5-year follow-up should be performed similarly to that in patients with gastric adenocarcinoma.
In conclusion, most gastric NETs are indolent and often show benign features. However, when a tumor is diagnosed as a NET, particularly when it is confirmed as a G2 or a higher-grade tumor, prompt endoscopic or surgical treatment should be performed to improve the patient’s prognosis.
No potential conflict of interest relevant to this article was reported.
Study conception and design: Y.K., B.A., S.M.H., J.H.L. Data acquisition: Y.K., B.A., K.D.C., J.H.Y., B.S.K., G.H.L., S.M.H., J.H.L. Data analysis and interpretation: Y.K., B.A., S.M.H., J.H.L. Manuscript writing: Y.K., B.A., S.M.H., J.H.L. Critical revision of the manuscript for important intellectual content: S.M.H., J.H.L. Approval of final manuscript: all authors.
Supplementary materials can be accessed at https://doi.org/10.5009/gnl220175.
Gut and Liver 2023; 17(6): 863-873
Published online November 15, 2023 https://doi.org/10.5009/gnl220175
Copyright © Gut and Liver.
Yuri Kim1 , Bokyung Ahn2 , Kee Don Choi1 , Beom-Su Kim3 , Jeong-Hwan Yook3 , Gin Hyug Lee1 , Seung-Mo Hong2 , Jeong Hoon Lee1
Departments of 1Gastroenterology, 2Pathology, and 3Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea
Correspondence to:Seung-Mo Hong
ORCID https://orcid.org/0000-0002-8888-6007
E-mail smhong28@gmail.com
Jeong Hoon Lee
ORCID https://orcid.org/0000-0002-0778-7585
E-mail jhlee.gi@amc.seoul.kr
Yuri Kim and Bokyung Ahn are contributed equally to this work as first authors.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background/Aims: Although gastric neuroendocrine tumors (NETs) are uncommon neoplasms, their prevalence is increasing. The clinical importance of the World Health Organization (WHO) classification of gastric NETs, compared with NETs in other organs, has been underestimated. This study aimed to systematically evaluate the clinical and pathologic characteristics of gastric NETs based on the 2019 WHO classification and to assess the survival outcomes of patients from a single-center with a long-term follow-up.
Methods: The medical records of 427 patients with gastric NETs who underwent endoscopic or surgical resection between January 2000 and March 2020 were retrospectively reviewed. All specimens were reclassified according to the 2019 WHO classification. The clinicopathologic characteristics, treatment, and oncologic outcomes of 139 gastric NETs were analyzed.
Results: The patients’ median age was 53.0 years (interquartile range [IQR], 46.0 to 63.0 years). The median follow-up period was 36.0 months (IQR, 15.0 to 63.0 months). Of the patients, 92, 44, and 3 had grades 1, 2, and 3 NETs, respectively. The mean tumor size significantly increased as the tumor grade increased (p=0.025). Patients with grades 2 and 3 gastric NETs more frequently had lymphovascular invasion (29.8% vs 10.9%, p=0.005) and deeper tissue invasion (8.5% vs 0%, p=0.012) than those with grade 1 tumors. The overall disease-specific survival rate was 100%. Two patients with grades 2-3 gastric NETs experienced extragastric recurrence.
Conclusions: Although gastric NETs have an excellent prognosis, grade 2 or grade 3 gastric NETs are associated with a larger size, deeper invasion, and extragastric recurrence, which require active treatment.
Keywords: Neuroendocrine tumor, Stomach, World Health Organization, Classification, Survival
Since Oberndorfer1 first described the term “karzinoide” in 1907, neuroendocrine tumors (NETs) have been consistently found in the different organs of the gastrointestinal (GI) tract, and their prevalence has been gradually increasing owing to the increasing number of patients who undergo routine medical examinations.1 Gastric NETs are rare types of tumors, with an overall incidence of 0.2 cases per 100,000 people and a prevalence of 0.6% to 2% of all GI NETs.2-5
The most common gastric NETs are enterochromaffin-like (ECL)-cell (histamine-producing) tumors.6,7 In 1993, Rindi et al.8 first classified gastric ECL-cell NETs into three subtypes based on the histology of the adjacent mucosa, presence or absence of antral G-cell hyperplasia and hypergastrinemia, and accompanying clinical conditions. Type I gastric ECL NETs are the most common subtype. These are multiple small tumors occurring in the body or fundus and are associated with hypergastrinemia, G-cell hyperplasia, and autoimmune chronic atrophic gastritis. Type II ECL NETs develop in the body and fundus, with hypertrophy of the adjacent mucosa, and measure <2 cm in size. These tumors occur in patients with multiple endocrine neoplasia type 1 or Zollinger-Ellison syndrome and are accompanied by multiple endocrine neoplasms in the parathyroid gland, pituitary gland, and pancreas.9 Lastly, type III ECL NETs are sporadic tumors measuring >2 cm in size. They present as a single mass in any part of the stomach and are not associated with hypergastrinemia, autoimmune chronic atrophic gastritis, multiple endocrine neoplasia type 1, or Zollinger-Ellison syndrome.7,9,10 This classification method is only used for gastric NETs and not for NETs in other organs, with no substantial change since it was introduced in 1993.
Unlike the tumors in other organs, stomach tumors were distinguished using not only the World Health Organization (WHO) classification system but also the classification methods according to types. The 2019 WHO classification of gastric NETs was updated recently to better stratify the patients’ prognosis according to the following grading system: grade 1 (G1), <2 mitoses/2 mm2 in 10 high-power fields (HPFs) at 40× magnification and an ocular field diameter of 0.5 mm, and Ki-67 labeling index (Ki67LI) of <3%; grade 2 (G2), 2–20 mitoses/2 mm2 or Ki67LI of 3% to 20%; and grade 3 (G3), ≥20 mitoses/2 mm2 or Ki67LI of >20%.11 Although a few previous studies investigating gastric NETs have been published, they only included a small number of cases, lack uniformity in classifying the cases using the previous WHO criteria, and lack clinicopathologic or survival data.12-14 Therefore, this study aimed to systematically evaluate the clinical characteristics and pathologic findings of gastric NETs subclassified according to the updated 2019 WHO classification in a large-scale cohort of patients.
This single-center, retrospective study was conducted in Asan Medical Center, Seoul, Korea. A total of 427 patients diagnosed with NETs, well-differentiated NETs, poorly differentiated NETs, well-differentiated neuroendocrine neoplasms, well-differentiated neuroendocrine carcinoma (NEC), poorly differentiated NEC, and carcinoid tumors between January 2000 and March 2020 were recruited for this study. Fig. 1 shows a flowchart demonstrating the inclusion and exclusion criteria used to screen the initial cohort. A total of 139 patients with well-differentiated gastric NETs were finally selected for subsequent analysis.
An endoscopic evaluation was performed to assess the characteristics of NETs (including the size, number, and location of tumors) and the surrounding gastric mucosa. The location of gastric NETs was divided into the upper, middle, lower, and multiregional parts of the stomach. The upper part included the fundus, cardia, and high body; the middle part included the mid and lower body; and the lower part included the gastric angle and antral portion. Atrophy of the gastric mucosa was evaluated and classified as either closed or open type according to the Kimura-Takemoto classification.15 Intragastric recurrence was classified as synchronous recurrence and metachronous recurrence. Synchronous gastric NETs were defined as new gastric NETs additionally diagnosed within 6 months after the previous diagnosis, while metachronous gastric NETs were defined as new lesions detected after 6 months had elapsed.16 Extragastric recurrence was defined as tumor recurrence in organs other than the stomach after endoscopic or surgical resection of gastric NETs.
All patients underwent endoscopic (n=116) or surgical (n=23) resection. Although there are no clear therapeutic criteria, endoscopic resection was considered first when the size of the lesion was ≤2 cm, with no evidence of extragastric metastasis, and when the lesion was localized in the mucosal or submucosal layer on endoscopic ultrasonography. In contrast, when the size of the lesion was relatively large, or was accompanied by a deep ulcer on the surface making endoscopic resection difficult, or was deeper than the submucosal layer in endoscopic ultrasonography, or when the patient desired surgery, surgical treatment was performed. The different endoscopic treatments included endoscopic submucosal dissection (ESD), endoscopic mucosal resection, and argon plasma coagulation. The endoscopic procedures were performed by six expert endoscopists. Each endoscopic procedure was carefully selected and performed at the discretion of the endoscopists and according to the characteristics of the gastric NETs. The surgical methods performed were wedge resection, distal gastrectomy, and total gastrectomy. With regard to the endoscopic treatments, the surgical methods and extent of resection were determined by the GI surgeons. Follow-up esophagogastroduodenoscopy (EGD) or abdominal computed tomography examinations were performed every 6 to 12 months for the first 2 years after the endoscopic or surgical treatment and annually thereafter.
All hematoxylin and eosin-stained slides were carefully re-evaluated by two pathologists who were blinded to the patients’ clinicopathologic information. The tumor grade was evaluated using mitotic counts/10 HPFs and Ki67LI according to 2019 WHO classification system.11 Briefly, G1 was assigned to tumors with <2 mitoses/10 HPFs or Ki67LI of <3%; G2, to tumors with 2–20 mitoses/10 HPFs or Ki67LI of 3% to 20%; and G3, to tumors with >20 mitoses/10 HPFs or Ki67LI of >20%.11 Each tumor slide was manually scanned under a ×10 objective, and the area with the greatest Ki-67 positivity (“hotspot”) was selected for image capturing and printing. If the hotspot included <500 cells (the recommended minimum count of tumor cells for accurate grading), ≥2 HPFs were additionally selected, if available, to include a sufficient number of tumor cells.11 MicroNETs were defined as nodules of ECL cells with a diameter ranging from 0.05 to 0.5 cm.17 Neuroendocrine dysplasia (micronodules of neuroendocrine cells measuring >150 µm and <0.5 mm in size), NECs (small cell carcinomas or large cell carcinomas with sheets or poorly formed trabeculae or nests of neuroendocrine cells), and mixed neuroendocrine and non-neuroendocrine neoplasms (neoplasms with morphologically recognizable neuroendocrine and non-neuroendocrine components, each constituting ≥30% of the tumor volume) were excluded before and after central pathology review.12,18 Lymphovascular and perineural invasions, marginal status, and lymph node and distant metastatic status were also evaluated. The tumor stage was evaluated based on the 8th edition of the American Joint Committee on Cancer Staging Manual.18
Thirty-one cases with no available Ki67LI data were selected for Ki-67 immunohistochemistry. Immunohistochemical staining of the tumor tissue samples expressing Ki-67 was performed using the BenchMark XT autoimmunostainer (Ventana Medical Systems, Tucson, AZ, USA) with the OptiView DAB Detection Kit (Ventana Medical Systems) according to the manufacturer’s instructions and using the reagents supplied along with the kit. Briefly, 4-µm sections were mounted on silanized charged slides, allowed to dry for 10 minutes at room temperature, and placed in an incubator for 20 minutes at 65°C. The sections were treated with heat induced epitope retrieval (CC1) for 64 minutes and incubated for 32 minutes with mouse anti-Ki-67 antibody (clone MIB1, dilution 1:200; DAKO, Glostrup, Denmark) in the autoimmunostainer.
The baseline characteristics of gastric NETs subclassified according to the tumor grade were analyzed. Quantitative data (e.g., patients’ age, follow-up duration, and tumor size) were expressed as medians with the interquartile range. Qualitative data (e.g., patients’ sex; symptoms; multiorgan involvement; and number, location, atrophy, and pathologic characteristics of tumor specimens) were expressed as proportions. The categorical variables were compared using the non-parametric Kruskal-Wallis test. Owing to the small number of G3 gastric NETs, G2 and G3 NETs were grouped together (G2-3) and were compared with G1 gastric NETs. The clinical characteristics and outcomes of patients with microNETs were separately analyzed by conducting a subgroup analysis. A probability (p) value of <0.05 was considered significant. All statistical analyses were performed using IBM SPSS Statistics for Windows version 24.0 (IBM Corp., Armonk, NY, USA) and Prism 9 for Windows version 9.3.1 (GraphPad Software Inc., San Diego, CA, USA).
The study was approved by the Institutional Review Board of Asan Medical Center (IRB number: 2020-0996). The requirement for informed consent was waived because the data were retrospectively collected. All procedures were performed in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the 1964 Declaration of Helsinki and its later versions.
The grades of all specimens were re-evaluated according to the updated 2019 WHO classification system through a centralized pathology review performed by expert pathologists; this resulted in changes in gastric NET grade in some cases. After the central pathology review, 139 patients were re-evaluated. The representative endoscopic and pathologic images of gastric NETs by tumor grade are depicted in Fig. 2.
The baseline characteristics of the patients are summarized in Table 1. Of the 139 patients, 92 (66.2%), 44 (31.7%), and three (2.1%) had G1, G2, and G3 NETs, respectively. Most patients were asymptomatic (62.0% of those with G1 NETs and 70.5% of those with G2 NETs), and gastric NETs were often incidentally detected during routine EGD examinations. However, all three patients with G3 NETs experienced abdominal pain prior to the diagnosis.
Table 1 . Patients’ Baseline Characteristics and Endoscopic Findings.
Characteristic | Total (n=139) | Grade 1 (n=92) | Grade 2 (n=44) | Grade 3 (n=3) | p-value* |
---|---|---|---|---|---|
Age, yr | 53.0 (46.0–63.0) | 53.5 (46.0–63.0) | 51.5 (43.3–60.0) | 59.0 (52.0–59.0) | 0.311 |
Follow-up period, mo | 36.0 (15.0–63.0) | 31.5 (13.5–61.8) | 42.5 (22.8–68.0) | 35.0 (15.0–35.0) | 0.481 |
Sex | 0.765 | ||||
Male | 74 (53.2) | 50 (54.3) | 23 (52.3) | 1 (33.3) | |
Female | 65 (46.8) | 42 (45.7) | 21 (47.7) | 2 (66.7) | |
Symptom | 0.077 | ||||
None | 88 (63.3) | 57 (62.0) | 31 (70.5) | 0 | |
Pain | 46 (33.1) | 33 (35.8) | 10 (22.7) | 3 (100.0) | |
Bleeding | 2 (1.4) | 0 | 2 (4.5) | 0 | |
Carcinoid symptom | 3 (2.2) | 2 (2.2) | 1 (2.3) | 0 | |
Multiorgan involvement | 6 (4.2) | 4 (4.3) | 2 (4.6) | 3 (100.0) | 0.931 |
Tumor size, mm | 10.0 (6.0–10.0) | 8.0 (5.0–10.0) | 10.0 (6.0–15.0) | 15.0 (10.0–15.0) | 0.025† |
Number | 0.558 | ||||
Single | 109 (78.4) | 73 (79.3) | 33 (75.0) | 3 (100.0) | |
Multifocal | 30 (21.6) | 19 (20.7) | 11 (25.0) | 0 | |
Location | 0.042† | ||||
Upper | 47 (33.8) | 32 (34.8) | 15 (34.1) | 0 | |
Middle | 71 (51.1) | 49 (53.2) | 22 (50.0) | 0 | |
Lower | 7 (5.0) | 3 (3.3) | 1 (2.3) | 3 (100.0) | |
Multiregional | 14 (10.1) | 8 (8.7) | 6 (13.6) | 0 | |
Atrophy | 0.588 | ||||
None | 23 (16.5) | 17 (18.5) | 5 (11.4) | 1 (33.3) | |
Closed type | 19 (13.7) | 13 (14.1) | 6 (13.6) | 0 | |
Open type | 97 (69.8) | 62 (67.4) | 33 (75.0) | 2 (66.7) | |
Peptic ulcer disease | 7 (5.0) | 3 (3.3) | 2 (4.5) | 2 (66.7) | <0.001† |
Data are presented as median (interquartile range) or number (%)..
*Statistical significance was assessed using the Kruskal-Wallis test; †p<0.05..
The endoscopic findings of each tumor grade are summarized in Table 1. The median macroscopic size of gastric NETs was significantly larger in patients with higher-grade tumors (G1 vs G2 vs G3, 8.0 mm vs 10.0 mm vs 15.0 mm, p=0.025). A total of 71 tumors (51.1%) were detected in the middle part of the stomach, with all G3 NETs located in the lower part (53.3%, 50.0%, and 0% for G1, G2, and G3, respectively, p=0.042). Peptic ulcers rarely occurred in patients with G1 and G2 NETs; however, two of the three G3 NETs were detected in the form of multiple ulcers (3.3% vs 4.5% vs 66.7%, p<0.001).
Owing to the small number of patients with G3 tumors, those with G2 and G3 tumors were merged and analyzed as one group. A total of 119 patients (85.6%) underwent at least one endoscopic procedure (Table 2, Supplementary Fig. 1). Eleven patients received repeated endoscopic procedures. Patients with a synchronous lesion size with a size or shape change (eight patients) or a metachronous lesion (one patient) and suspected recurrence at the previous resected site (one patient) underwent repeated endoscopic treatment. The number of patients who underwent surgical treatment was significantly higher in the G2-3 NET group than in the G1 NET group (10.9% for G1 vs 38.3% for G2-3, p<0.001). Lymph node metastasis was confirmed in six of 22 patients who underwent surgical treatment (one patient with G1 NET and five with G2 NETs). Lymphovascular invasion was more frequently observed in patients with G2-3 NETs than in those with G1 NETs (10.9% for G1 vs 29.8% for G2-3, p=0.005). However, no significant differences were observed in perineural invasion and resection margin status. Most of the patients had gastric NETs in the mucosal or submucosal layer, whereas serosal invasion was confirmed by pathologic review in four patients with G2 NETs (2.9%). Additional chemotherapy treatment was more frequently administered in patients with G2-3 NETs than in those with G1 NETs (1.1% for G1 vs 10.7% for G2-3, p=0.009). When the treatment results were analyzed according to the resection method, the specimens obtained through surgical resection had a larger size, higher pathological grade and negative resection margin rates, lower rates for lymphovascular invasion and perineural invasion, and deeper invasion depth (Supplementary Table 1).
Table 2 . Comparison of Treatment and Oncologic Outcomes.
Variable | Total (n=139) | Grade 1 (n=92) | Grade 2-3 (n=47) | p-value* |
---|---|---|---|---|
Endoscopic treatment | 0.058 | |||
None | 20 (14.4) | 8 (8.7) | 12 (25.5) | |
1 Time | 109 (78.4) | 78 (84.8) | 31 (66.0) | |
2 Times | 8 (5.8) | 4 (4.3) | 4 (8.5) | |
3 Times | 2 (1.4) | 2 (2.2) | 0 | |
Surgery | <0.001† | |||
None | 111 (79.9) | 82 (89.1) | 29 (61.7) | |
Wedge resection | 11 (7.9) | 5 (5.4) | 6 (12.8) | |
Subtotal gastrectomy | 5 (3.6) | 1 (1.1) | 4 (8.5) | |
Total gastrectomy | 12 (8.6) | 4 (4.3) | 8 (17.0) | |
Pathologic size, mm | 8.0 (6.0–12.0) | 7.0 (4.3–9.0) | 11.0 (6.0–15.0) | <0.001† |
Lymphovascular invasion | 0.005† | |||
Negative | 115 (82.7) | 82 (89.1) | 33 (70.2) | |
Positive | 24 (17.3) | 10 (10.9) | 14 (29.8) | |
Perineural invasion | 0.162 | |||
Negative | 138 (99.3) | 92 (100.0) | 46 (97.9) | |
Positive | 1 (0.7) | 0 | 1 (2.1) | |
Resection margin | 0.210 | |||
Negative | 117 (84.2) | 80 (87.0) | 37 (78.7) | |
Positive | 22 (15.8) | 12 (13.0) | 10 (21.3) | |
Invasion depth | 0.012† | |||
Mucosa | 30 (21.6) | 24 (26.1) | 6 (12.8) | |
Submucosa | 105 (75.5) | 68 (73.9) | 37 (78.7) | |
Muscularis propria | 0 | 0 | 0 | |
Serosa | 4 (2.9) | 0 | 4 (8.5) | |
Chemotherapy | 0.009† | |||
None | 133 (95.7) | 91 (98.9) | 42 (89.4) | |
Adjuvant | 3 (2.2) | 1 (1.1) | 2 (4.2) | |
Palliative | 3 (2.2) | 0 | 3 (6.4) | |
Intragastric recurrence | 0.966 | |||
None | 107 (77.0) | 71 (77.2) | 36 (76.6) | |
Synchronous | 31 (22.3) | 20 (21.7) | 11 (23.4) | |
Metachronous | 1 (0.7) | 1 (1.1) | 0 | |
Extragastric recurrence | 0.047† | |||
None | 137 (98.6) | 92 (100.0) | 45 (95.7) | |
Recurred | 2 (1.4) | 0 | 2 (4.3) | |
All-cause mortality | 0.754 | |||
Death | 9 (6.5) | 5 (5.4) | 4 (8.5) | |
Survived | 130 (93.5) | 87 (94.6) | 43 (91.5) |
Data are presented as number (%) or median (interquartile range)..
*Statistical significance was assessed using the Kruskal-Wallis test; †p<0.05..
In patients who underwent endoscopic procedures, 15 patients were lymphovascular invasion positive. Among them, five patients underwent additional surgical treatment (two wedge resections, two subtotal gastrectomy, and one total gastrectomy). This decision was based on the depth of submucosal invasion, positive resection margin in addition to lymphovascular invasion positivity, or the patient’s desire for additional surgical treatment. There was no lesion recurrence in any of the patients, regardless of whether they underwent additional surgery.
There was one patient who had metachronous gastric NET. A 44-year-old female underwent ESD for a 12-mm-sized submucosal tumor-like NET on the posterior wall of the mid body. Seven years after the first procedure, a 3-mm sized G1 NET was newly detected in the posterior wall of the high body, and additional endoscopic mucosal resection was performed. After two procedures, no additional recurrence was found during regular follow-up endoscopy.
Two patients (1.4%) experienced extragastric recurrence (Table 2). One patient was a 40-year-old woman who underwent total gastrectomy for a 30-mm gastric NET in the mid body, which was histologically confirmed as a G2 tumor. Four months after surgery, peritoneal carcinomatosis was detected on follow-up abdominal computed tomography; hence, palliative irinotecan chemotherapy was administered. The other patient was a 52-year-old woman who experienced epigastric discomfort and had a 10-mm NET in the antrum, which was detected during EGD and histologically confirmed as a G3 tumor with no lymphovascular or perineural invasion and with a clear resection margin. Follow-up EGD performed 32 months after the initial treatment showed G3 NET recurrence at the post-ESD site, with multiple bones and liver metastases identified on abdominal computed tomography images. The patient was treated with lanreotide and everolimus and underwent combination chemotherapy with cisplatin and fluorouracil.
Nine patients (6.5%) died during the study (five patients with G1 NETs and four patients with G2 NETs), with no significant survival difference between the groups (94.6% for G1 vs 91.5% for G2-3, p=0.754). None of the nine deaths were due to the progression of gastric NETs; six patients died of progression of other underlying diseases, while three patients died of unknown causes (Supplementary Table 2).
Of the 21 patients with microNETs (Figs 3-5), most had G1 NETs (16 cases, 76%) (Figs 4 and 5), whereas four patients had G2 NETs despite the small tumor size (Fig. 3). No significant differences were found in the clinical characteristics between the two groups. Five patients (23.8%) experienced intragastric recurrences during the observation period; however, none had extragastric recurrence (Supplementary Table 3).
Our study aimed to systematically re-evaluate the clinicopathologic characteristics and survival outcomes of patients with gastric NETs subclassified according to the updated 2019 WHO classification system. Previous three-tiered classifications of gastric NETs, unlike those of NETs in other organs, were mainly based on the presence or absence of atrophy and hypergastrinemia. However, such classification methods were limited by their ambiguity or the practical difficulties in conducting tests (e.g., gastrin test) in the clinical setting. Therefore, to compare the clinical characteristics and prognosis based on the pathologic classification alone, similar to the classification of NETs in other organs, this study was conducted after excluding type-related factors. In the present study, no significant differences were found in the patients’ age and sex among the tumor grades; however, a higher grade was found to be correlated with a larger tumor, higher rate of lymphovascular invasion, and a greater depth of invasion, even after excluding the factors related to the types of gastric NETs. Considering these results, the 2019 WHO classification alone may be sufficient in predicting the clinical features and prognosis of gastric NETs. Because our cohort consisted of patients diagnosed with gastric NETs within the last 20 years, when various terminologies were used with inaccurate tumor grade, lack of Ki-67 immunohistochemistry or inaccurate Ki67LI owing to “eyeballing” measurements, many of the patients were reclassified after a centralized review. In the process, cases that did not meet the current diagnostic criteria for gastric NETs were excluded from the study, thus increasing the accuracy of our results.
The prevalence of NETs has been gradually increasing owing to an increase in the number of patients undergoing routine medical evaluations.1,19 According to a number of studies the age-adjusted incidence of GI NETs in 1973 was 1.09 per 100,000, which significantly increased to 5.25 per 100,000 in 2004 and 6.98 per 100,000 in 2012 in the United States.20,21 The NETs that commonly occur in the GI tract are small-intestinal NETs (30.8%), followed by rectal (26.3%), colonic (17.6%), and pancreatic (12.1%) NETs. By contrast, gastric NETs are relatively infrequent (8.9%).22,23 Unlike the NETs occurring in other GI organs, which are classified solely based on pathologic grading, gastric NETs have been categorized according to the presence or absence of hypergastrinemia or atrophy.9 However, these classification methods are often difficult to apply in real-world clinical settings, especially in primary clinics, owing to the difficulties in measuring the serum gastrin levels and objectively evaluating gastric atrophy. Because gastrin is released by G cells in response to the stimulation of gastrin-releasing peptides in the stomach, and a person’s food intake or prolonged use of proton pump inhibitors can lead to fluctuations in serum gastrin levels, the results of the serum gastrin test should be interpreted with caution.24-26 Even if the gastrin levels and atrophy are assessed, several patients do not meet the criteria for ECL NET subtyping. Therefore, applying the conventional ECL-cell NET subtyping criteria is often ambiguous. In addition to this subtyping, the WHO classification of NETs based on the Ki-67 index and mitotic status was defined in 2010 and was recently revised in 2019.11,27 Ki-67 index was added to the classification criteria after Garcia-Carbonero et al.28 found that Ki-67 is an independent predictor of survival. However, only a few studies have determined the prognostic significance of classifying gastric NETs according to the 2019 WHO classification system, compared with ECL-cell NET typing; thus, its importance is relatively underestimated.
In this study, all pathologic specimens, including those diagnosed before 2010, were centrally reviewed and analyzed together; most gastric NETs were G1 tumors (66.2%) of <10 mm in size and developed in the upper part of the stomach. These characteristics are consistent with the features of type I ECL NETs, which account for 70% to 80% of all NETs (i.e., relatively small in size and often found in the fundus and body of the stomach).14,29 Although the all-cause mortality rate was reported to be 6.5%, the disease-specific mortality rate was 0%, and the number of patients with recurrent gastric NETs was extremely small. Thus, gastric NETs are considered to have an excellent overall prognosis. However, as the tumor grade increased, the gastric NETs showed a gradual increase in size and a greater association with lymphovascular invasion after endoscopic or surgical resection. Therefore, in patients with G2 or G3 gastric NETs, close observation is necessary owing to their higher risk of developing local recurrence after undergoing endoscopic or surgical treatment compared with those with G1 NETs. Clinical data on the recurrence rate and average recurrence period of gastric NETs were lacking, with no consensus on the follow-up period after treatment. Of the 139 patients, one patient with a G2 gastric NET experienced recurrence 4 months after surgery, while one patient with a G3 gastric NET experienced recurrence at 32 months after ESD, with a median follow-up of 36 months. Considering these findings, these patients should be closely observed for 5 years, which is similar to that of patients with gastric adenocarcinoma. In this regard, guidelines based on the results of further studies should be established.
All patients included in this study underwent surgery or endoscopic treatment, and those who had undergone endoscopic forceps biopsy alone were excluded owing to the extremely small specimen size. Because NETs are indolent and have a lower malignancy risk compared with other GI tumors, the National Comprehensive Cancer Network and European Neuroendocrine Tumor Society guidelines only recommend surveillance if the lesion is confined to the mucosal or submucosal layer or <10 mm in size.30-34 However, our study showed that even in patients with microNETs, unlike those with neuroendocrine microadenomas in the pancreas, higher-grade tumors (e.g., G2) may be left undetected.35,36 Furthermore, some experts suggested that it may be better to remove small gastric NETs using endoscopic methods, such as polypectomy, endoscopic mucosal resection, and ESD.37-40 Considering the results of these studies, if a gastric NET is diagnosed, active treatment using endoscopic or surgical methods should be considered, even for those with small gastric NETs.
This study has several limitations. First, the results may be limited by the retrospective and single-center study design. In addition, as patients with small gastric NETs who only underwent observation, those whose tumors were removed by forceps biopsy, and those without proper endoscopic or surgical specimens were excluded, selection bias may be present. Third, the number of patients with G3 NETs was relatively small, which might have confounded the analysis. However, three patients (2.1%) had G3 gastric NETs; this prevalence rate is consistent with that reported in previous studies, although the data on G3 NETs are limited owing to their recent distinction from NECs in 2017.41-43 In the PRONET study of neuroendocrine neoplasms in the lungs and digestive system, 21 patients (2.7%) had G3 NETs among 778 patients with GI neuroendocrine neoplasms, which was similar to the prevalence of G3 NETs in our study.41-43 Although the PRONET study differs from our study in that it examined all types of NETs, it provides evidence that G3 NETs are rare. Therefore, further studies with a larger number of patients with gastric NETs, including those who underwent observation or endoscopic biopsy alone, are needed to obtain more accurate results, including those of patients with G3 NETs. Finally, as two patients experienced recurrence at 4 and 32 months after the initial treatment of gastric NETs, and the median follow-up period of the overall patients was 36 months, the follow-up period in this study was relatively short. As mentioned above, for accurate observation, a 5-year follow-up should be performed similarly to that in patients with gastric adenocarcinoma.
In conclusion, most gastric NETs are indolent and often show benign features. However, when a tumor is diagnosed as a NET, particularly when it is confirmed as a G2 or a higher-grade tumor, prompt endoscopic or surgical treatment should be performed to improve the patient’s prognosis.
No potential conflict of interest relevant to this article was reported.
Study conception and design: Y.K., B.A., S.M.H., J.H.L. Data acquisition: Y.K., B.A., K.D.C., J.H.Y., B.S.K., G.H.L., S.M.H., J.H.L. Data analysis and interpretation: Y.K., B.A., S.M.H., J.H.L. Manuscript writing: Y.K., B.A., S.M.H., J.H.L. Critical revision of the manuscript for important intellectual content: S.M.H., J.H.L. Approval of final manuscript: all authors.
Supplementary materials can be accessed at https://doi.org/10.5009/gnl220175.
Table 1 Patients’ Baseline Characteristics and Endoscopic Findings
Characteristic | Total (n=139) | Grade 1 (n=92) | Grade 2 (n=44) | Grade 3 (n=3) | p-value* |
---|---|---|---|---|---|
Age, yr | 53.0 (46.0–63.0) | 53.5 (46.0–63.0) | 51.5 (43.3–60.0) | 59.0 (52.0–59.0) | 0.311 |
Follow-up period, mo | 36.0 (15.0–63.0) | 31.5 (13.5–61.8) | 42.5 (22.8–68.0) | 35.0 (15.0–35.0) | 0.481 |
Sex | 0.765 | ||||
Male | 74 (53.2) | 50 (54.3) | 23 (52.3) | 1 (33.3) | |
Female | 65 (46.8) | 42 (45.7) | 21 (47.7) | 2 (66.7) | |
Symptom | 0.077 | ||||
None | 88 (63.3) | 57 (62.0) | 31 (70.5) | 0 | |
Pain | 46 (33.1) | 33 (35.8) | 10 (22.7) | 3 (100.0) | |
Bleeding | 2 (1.4) | 0 | 2 (4.5) | 0 | |
Carcinoid symptom | 3 (2.2) | 2 (2.2) | 1 (2.3) | 0 | |
Multiorgan involvement | 6 (4.2) | 4 (4.3) | 2 (4.6) | 3 (100.0) | 0.931 |
Tumor size, mm | 10.0 (6.0–10.0) | 8.0 (5.0–10.0) | 10.0 (6.0–15.0) | 15.0 (10.0–15.0) | 0.025† |
Number | 0.558 | ||||
Single | 109 (78.4) | 73 (79.3) | 33 (75.0) | 3 (100.0) | |
Multifocal | 30 (21.6) | 19 (20.7) | 11 (25.0) | 0 | |
Location | 0.042† | ||||
Upper | 47 (33.8) | 32 (34.8) | 15 (34.1) | 0 | |
Middle | 71 (51.1) | 49 (53.2) | 22 (50.0) | 0 | |
Lower | 7 (5.0) | 3 (3.3) | 1 (2.3) | 3 (100.0) | |
Multiregional | 14 (10.1) | 8 (8.7) | 6 (13.6) | 0 | |
Atrophy | 0.588 | ||||
None | 23 (16.5) | 17 (18.5) | 5 (11.4) | 1 (33.3) | |
Closed type | 19 (13.7) | 13 (14.1) | 6 (13.6) | 0 | |
Open type | 97 (69.8) | 62 (67.4) | 33 (75.0) | 2 (66.7) | |
Peptic ulcer disease | 7 (5.0) | 3 (3.3) | 2 (4.5) | 2 (66.7) | <0.001† |
Data are presented as median (interquartile range) or number (%).
*Statistical significance was assessed using the Kruskal-Wallis test; †p<0.05.
Table 2 Comparison of Treatment and Oncologic Outcomes
Variable | Total (n=139) | Grade 1 (n=92) | Grade 2-3 (n=47) | p-value* |
---|---|---|---|---|
Endoscopic treatment | 0.058 | |||
None | 20 (14.4) | 8 (8.7) | 12 (25.5) | |
1 Time | 109 (78.4) | 78 (84.8) | 31 (66.0) | |
2 Times | 8 (5.8) | 4 (4.3) | 4 (8.5) | |
3 Times | 2 (1.4) | 2 (2.2) | 0 | |
Surgery | <0.001† | |||
None | 111 (79.9) | 82 (89.1) | 29 (61.7) | |
Wedge resection | 11 (7.9) | 5 (5.4) | 6 (12.8) | |
Subtotal gastrectomy | 5 (3.6) | 1 (1.1) | 4 (8.5) | |
Total gastrectomy | 12 (8.6) | 4 (4.3) | 8 (17.0) | |
Pathologic size, mm | 8.0 (6.0–12.0) | 7.0 (4.3–9.0) | 11.0 (6.0–15.0) | <0.001† |
Lymphovascular invasion | 0.005† | |||
Negative | 115 (82.7) | 82 (89.1) | 33 (70.2) | |
Positive | 24 (17.3) | 10 (10.9) | 14 (29.8) | |
Perineural invasion | 0.162 | |||
Negative | 138 (99.3) | 92 (100.0) | 46 (97.9) | |
Positive | 1 (0.7) | 0 | 1 (2.1) | |
Resection margin | 0.210 | |||
Negative | 117 (84.2) | 80 (87.0) | 37 (78.7) | |
Positive | 22 (15.8) | 12 (13.0) | 10 (21.3) | |
Invasion depth | 0.012† | |||
Mucosa | 30 (21.6) | 24 (26.1) | 6 (12.8) | |
Submucosa | 105 (75.5) | 68 (73.9) | 37 (78.7) | |
Muscularis propria | 0 | 0 | 0 | |
Serosa | 4 (2.9) | 0 | 4 (8.5) | |
Chemotherapy | 0.009† | |||
None | 133 (95.7) | 91 (98.9) | 42 (89.4) | |
Adjuvant | 3 (2.2) | 1 (1.1) | 2 (4.2) | |
Palliative | 3 (2.2) | 0 | 3 (6.4) | |
Intragastric recurrence | 0.966 | |||
None | 107 (77.0) | 71 (77.2) | 36 (76.6) | |
Synchronous | 31 (22.3) | 20 (21.7) | 11 (23.4) | |
Metachronous | 1 (0.7) | 1 (1.1) | 0 | |
Extragastric recurrence | 0.047† | |||
None | 137 (98.6) | 92 (100.0) | 45 (95.7) | |
Recurred | 2 (1.4) | 0 | 2 (4.3) | |
All-cause mortality | 0.754 | |||
Death | 9 (6.5) | 5 (5.4) | 4 (8.5) | |
Survived | 130 (93.5) | 87 (94.6) | 43 (91.5) |
Data are presented as number (%) or median (interquartile range).
*Statistical significance was assessed using the Kruskal-Wallis test; †p<0.05.