1.2.1 Pathology  1.2.1 病理学

Gastric neuroendocrine tumours (gNETs) are different from other gastrointestinal NETs as in the stomach a clear aetiology is emerging where gastric NETs are induced in the context of hyperplasia of enterochromaffin-like (ECL) cells. In 1993 an aetiological framework for gastric NETs was reported,² where gastric NETs were either found in a setting of ECL cell hyperplasia associated with high gastrin levels due to autoimmune fundic atrophic gastritis (type I NETs), high gastrin levels due to gastrinoma (type II NETs) or without ECL hyperplasia (type III NETs). More recently, general atrophy in chronic H. pylori associated gastritis,³ functional failure of parietal cells due to mutations⁴ or proton pump inhibitor (PPI) use⁵ have also been reported to be associated with NETs, some with a background of ECL cell hyperplasia but in cases of proton pump inhibition also associated with parietal cell hyperplasia. These newly described gastric NETs are similar to the originally described type I gNETs as they are secondary to failed acid secretion. The secondary (type I and type II) gNETs have a more indolent course^{2, 5} and can be reversible if the cause can be impeded. It is therefore important to recognise the associated gastric mucosal changes.⁶
胃神经内分泌肿瘤（gNETs）与其他胃肠道 NETs 不同，因为在胃部已明确其发病机制——胃 NETs 是在肠嗜铬样（ECL）细胞增生的背景下诱发的。1993 年提出的胃 NETs 病因学框架将胃 NETs 分为三类：与自身免疫性胃底萎缩性胃炎导致高胃泌素血症相关的 ECL 细胞增生型（I 型 NETs）、胃泌素瘤导致高胃泌素血症型（II 型 NETs）以及无 ECL 细胞增生型（III 型 NETs）。最新研究发现，慢性幽门螺杆菌相关性胃炎的整体萎缩、壁细胞功能缺陷（由基因突变或质子泵抑制剂（PPI）使用引起）也与 NETs 相关，部分病例存在 ECL 细胞增生背景，而质子泵抑制剂使用相关病例还伴有壁细胞增生。这些新发现的胃 NETs 与经典 I 型 gNETs 相似，均继发于胃酸分泌障碍。继发性（I 型和 II 型）gNETs 病程更为惰性，若阻断病因可实现病情逆转。 因此，识别伴随的胃黏膜变化至关重要。 ⁶

Adequate tumour sampling is essential for grading NETs as prescribed in the World Health Organization (WHO) classification of gastrointestinal tumours of 2019.⁷ As NETs are usually located deep in the mucosa (not at the surface) attention should be paid to this when taking biopsies.
根据 2019 年世界卫生组织（WHO）胃肠道肿瘤分类标准，充分的肿瘤取样对于神经内分泌肿瘤（NETs）的分级至关重要。 ⁷ 由于 NETs 通常位于黏膜深层（而非表面），活检时需特别注意这一点。

Gastric NETs are graded using the same criteria as for all gastrointestinal sites:⁸
胃神经内分泌肿瘤（Gastric NETs）的分级标准与所有胃肠道部位相同： ⁸

G1: mitotic count <2 in 2 mm² and/or Ki-67 index <3% in hotspot of at least 500 cells.
G1 级：每 2 平方毫米有丝分裂计数<2 个 ² ，和/或热点区域至少 500 个细胞中 Ki-67 指数<3%。

G2: mitotic count between 2 and 20 in 2 mm² and/or Ki-67 index between 3% and 20% in hotspot of at least 500 cells.
G2 级：每 2 平方毫米有丝分裂计数 2-20 个 ² ，和/或热点区域至少 500 个细胞中 Ki-67 指数 3%-20%。

G3: mitotic count >20 in 2 mm² and/or Ki-67 index >20% in hotspot of at least 500 cells.
G3 级：每 2 平方毫米有丝分裂计数>20 个 ² ，和/或热点区域至少 500 个细胞中 Ki-67 指数>20%。

1.2.2 Endoscopy  1.2.2 内镜检查

The primary diagnostic tool for gNETs is oesophagogastroduodenoscopy (OGD). For the assessment of gastric NETs, not only representative biopsies of the tumour but also biopsies of the surrounding mucosa and other parts of the stomach (antrum and body-fundus) are essential as these give clues on the nature/aetiology and, with this, the prognosis and treatment of the NET. In addition, sampling of gastric juice for pH measurement is strongly encouraged. Furthermore, virtual chromoendoscopy (narrow-band imaging, blue light imaging, i-scan) may be useful to provide additional information to identify field changes or precancerous lesions in underlying atrophic mucosa.⁹ In addition to OGD, endoscopic ultrasound (EUS) should be performed in all lesions >1 cm regardless of type and in all lesions of type III regardless of size, unless large or metastatic lesions are found. Reporting should include size of the lesion infiltration depth and assessment of local lymph nodes.
胃神经内分泌肿瘤（gNETs）的主要诊断工具是食管胃十二指肠镜检查（OGD）。对于胃 NETs 的评估，不仅需要获取肿瘤的代表性活检标本，还必须对周围黏膜及胃的其他部位（如胃窦和胃体-胃底）进行活检，因为这些活检结果能提供关于肿瘤性质/病因的线索，进而影响其预后和治疗方案。此外，强烈建议采集胃液进行 pH 值测定。虚拟染色内镜技术（窄带成像、蓝光成像、i-scan）可能有助于提供额外信息，以识别潜在萎缩性黏膜中的区域改变或癌前病变。 ⁹ 除 OGD 外，所有直径>1 cm 的病灶（无论类型）以及所有 III 型病灶（无论大小）均应进行超声内镜（EUS）检查，除非发现病灶较大或存在转移。报告内容应包括病灶大小、浸润深度及局部淋巴结评估。

1.2.3 Imaging and laboratory examinations
1.2.3 影像学与实验室检查

Biomarker assessment should include fasting gastrin and chromogranin A (CgA). Whereas gastrin may be useful to differentiate between different types (elevated in type I and type II, normal in type III), CgA is only useful as a tumour marker in type III gNETs. In the presence of ongoing PPI therapy gastrin and CgA are not diagnostic and are difficult to interpret. In general, caution should be used when evaluating CgA values, considering the different diagnostic accuracy of available assays. Parietal cell and intrinsic factor antibodies, vitamin B12 and thyroid function parameters should be measured in suspected cases of autoimmune gastritis.
生物标志物检测应包括空腹胃泌素和嗜铬粒蛋白 A（CgA）。胃泌素有助于区分不同类型（I 型和 II 型升高，III 型正常），而 CgA 仅作为 III 型 gNET 的肿瘤标志物有效。在持续使用质子泵抑制剂治疗的情况下，胃泌素和 CgA 无诊断价值且难以解读。总体而言，评估 CgA 值时需谨慎，考虑不同检测方法的诊断准确性差异。疑似自身免疫性胃炎病例应检测壁细胞抗体、内因子抗体、维生素 B12 及甲状腺功能参数。

Whereas type I gNETs do not usually require additional cross-sectional imaging, type III gNETs should undergo cross-sectional imaging – liver contrast-enhanced magnetic resonance imaging (MRI) and/or contrast-enhanced thoracoabdominal computed tomography (CT), with previous ingestion of water immediately before examination, to fill stomach and duodenum and better depict small enhancing lesion in the arterial phase, and functional imaging, preferentially somatostatin receptor PET/CT (⁶⁸Ga-SSA-PET-CT).
I 型胃神经内分泌肿瘤通常无需额外进行横断面成像检查，而 III 型胃神经内分泌肿瘤则应接受横断面成像——肝脏对比增强磁共振成像（MRI）和/或对比增强胸腹部计算机断层扫描（CT），检查前立即饮水以充盈胃和十二指肠，从而在动脉期更好地显示小强化病灶，同时进行功能成像，优先选择生长抑素受体 PET/CT（ ⁶⁸ 镓-SSA-PET-CT）。

These should also be considered in cases of type I gNETs with metastases on EUS or high-risk features of metastases (G2, lymphatic or vascular invasion, size ≥1 cm). Recently, ⁶⁸Ga-SSA-PET-CT has been suggested in selected cases with tumour diameter >1 cm, G2 tumours (perhaps Ki-67 > 10 but exact cutoff not determined), or R1 margins after endoscopic resection.¹⁰
对于 I 型胃神经内分泌肿瘤，若超声内镜（EUS）发现转移灶或存在转移高风险特征（G2 分级、淋巴或血管浸润、肿瘤大小≥1 厘米），也应考虑上述检查。近期有建议指出，对于肿瘤直径>1 厘米、G2 分级（可能 Ki-67>10 但具体临界值未定）或内镜切除后切缘阳性（R1）的特定病例，可采用 ⁶⁸ 镓-SSA-PET-CT 检查。 ¹⁰

Prognosis depends on type, size, grading, histological risk factors (e.g., lymphovascular invasion, perineural invasion), completeness of resection and initial tumour stage.
预后取决于肿瘤类型、大小、分级、组织学危险因素（如淋巴血管侵犯、神经周围侵犯）、切除完整性及初始肿瘤分期。

1.3.1 Introduction  1.3.1 引言

Treatment options for type I gNETs include surveillance, endoscopic resection, somatostatin analogues (SSA) and eventually surgery. As prognosis is usually quite favourable and tumour growth slow and risk of metastases is below 1% in tumours <10 mm, all NET <1 cm can be observed without any need for intervention.^{11, 12} The ideal schedule for observational OGD is not defined and clinical practice ranges from every 6 months to every 2 years. Planning the first follow-up after 6 months and then every 12 months is the most widely used approach. A regular rebiopsy of lesions is not necessary unless atypical features (e.g., ulceration, erosion, pitting) appear, suggesting invasive progression of the lesion. The management of type I gNETs is summarised in Figure 1.
对于 I 型胃神经内分泌肿瘤（gNETs）的治疗选择包括监测、内镜下切除、生长抑素类似物（SSA）及必要时的手术治疗。由于此类肿瘤预后通常较好，生长缓慢且小于 10 毫米的肿瘤转移风险低于 1%，所有直径小于 1 厘米的 NET 均可采取观察策略而无需干预。 ^{11, 12} 理想的观察性胃镜检查间隔尚未明确，临床实践中从每 6 个月到每 2 年不等。首次随访安排在 6 个月后，之后每 12 个月复查一次是最普遍采用的方式。除非出现提示病变侵袭性进展的非典型特征（如溃疡、糜烂、凹陷），否则无需定期重复活检。I 型 gNETs 的管理流程总结见图 1。

1.3.2 Q1 – What is the role of endoscopic resection for type I gNETs? Is additional treatment required in cases of R1 endoscopic resection?
1.3.2 问题 1 – 内镜切除术在 I 型 gNETs 中的作用是什么？对于 R1 内镜切除术后的病例，是否需要额外治疗？

Endoscopic resection should be considered for type I gNETs larger than 1 cm and for those lesions demonstrating increased Ki-67 as these features are associated with an increased risk of metastases and progression.^{11, 13} Many small type I gNETs are G1 or low G2 (Ki-67 < 10%);¹⁴ however, a Ki-67 cutoff has not been defined to determined when excision should be performed. Endoscopic ultrasonography to determine depth of local invasion and assessment of local lymph nodes should be performed in lesions >1 cm and in smaller tumours with “high” G2 grade (cutoff not established) before any resection. No other imaging modality is required at this stage. The resection technique of choice depends on size and position of the lesion, invasion depth and local experience. Endoscopic mucosal resection (EMR), endoscopic submucosal dissection (ESD) and full-thickness resection (FTR) are all associated with a low risk profile for both bleeding and perforation. It is not clear which is the best endoscopic technique to be used for achieving a complete R0 resection.¹⁵ However, resection by ESD or FTR achieve higher rates of R0 resection, but no randomised trials have compared all techniques head-to-head.^{16, 17} For R1 resection status a re-resection in a “step-up” approach could be recommended (e.g., ESD after EMR, FTR after ESD), although data showing risk of local recurrence after R1 resection are scarce,¹⁸ particularly for smaller tumours which may be managed by a non-interventional endoscopic surveillance after initial R1 resection.¹⁹
对于大于 1 厘米的 I 型胃神经内分泌肿瘤（gNETs）以及 Ki-67 升高的病灶，应考虑进行内镜下切除，因为这些特征与转移和进展风险增加相关。 ^{11, 13} 许多小型 I 型 gNETs 为 G1 或低级别 G2（Ki-67 < 10%）； ¹⁴ 然而，目前尚未确定 Ki-67 的临界值以决定何时应进行切除。对于大于 1 厘米的病灶以及“高级别”G2（临界值未确定）的较小肿瘤，在切除前应进行超声内镜检查以确定局部浸润深度并评估局部淋巴结情况。此阶段无需其他影像学检查。切除技术的选择取决于病灶的大小、位置、浸润深度及当地经验。内镜下黏膜切除术（EMR）、内镜黏膜下剥离术（ESD）和全层切除术（FTR）均具有较低的出血和穿孔风险。目前尚不清楚哪种内镜技术最适合实现完整的 R0 切除。 ¹⁵ 然而，内镜黏膜下剥离术（ESD）或全层切除术（FTR）可实现更高的 R0 切除率，但目前尚无随机试验直接比较所有技术。 ^{16, 17} 对于 R1 切除状态，可推荐采用“逐步升级”方法进行再切除（例如 EMR 后行 ESD，ESD 后行 FTR），尽管显示 R1 切除后局部复发风险的数据有限， ¹⁸ 特别是对于较小的肿瘤，初次 R1 切除后可能通过非干预性内镜监测进行处理。 ¹⁹

1.3.3 Q2 – When should somatostatin analogues be used to treat type I gNETs?
1.3.3 Q2 – 何时应使用生长抑素类似物治疗 I 型胃神经内分泌肿瘤？

Type I gNETs usually express somatostatin receptors. Thus, therapy with SSA can be initiated in metastasised patients or patients not amenable to endoscopic or surgical resection (e.g., difficult location, old age, comorbidities). SSA therapy is associated with a high complete response rate of 25%–100%, but relapse is frequently observed after discontinuation of therapy.²⁰ Thus, continuous therapy would be the appropriate approach. Another reason for starting SSA therapy could be large tumour size or the requirement of repeated endoscopic resection for progressive lesions, although data evaluating this approach are lacking. Another therapeutic approach could be direct inhibition of gastrin signalling in tumour cells. Indeed, the gastrin receptor inhibitor netazepide has been evaluated in a proof-of-concept phase II trial with 16 patients showing a complete response rate in 30% of patients.²¹ Like SSA treatment, tumour relapse after discontinuation was observed in all patients. The clinical benefit of netazepide needs to be evaluated in larger randomised clinical trials before any recommendation for its use can be formulated.
I 型胃神经内分泌肿瘤通常表达生长抑素受体。因此，对于发生转移或无法进行内镜或手术切除（如位置困难、高龄、合并症）的患者，可启动生长抑素类似物（SSA）治疗。SSA 治疗与 25%-100%的高完全缓解率相关，但停药后常观察到复发。 ²⁰ 因此，持续治疗可能是更合适的方法。启动 SSA 治疗的其他原因可能包括肿瘤体积较大或需要对进展性病变进行重复内镜切除，尽管目前缺乏评估这种方法的数据。另一种治疗方法可能是直接抑制肿瘤细胞中的胃泌素信号传导。事实上，胃泌素受体抑制剂 netazepide 已在一项概念验证性 II 期试验中进行了评估，16 名患者中有 30%显示完全缓解。 ²¹ 与 SSA 治疗类似，所有患者在停药后均观察到肿瘤复发。在提出任何使用建议之前，需要在更大规模的随机临床试验中评估 netazepide 的临床益处。

1.3.4 Q3 – What is the recommended follow-up schedule in type I gNETs?
1.3.4 问题 3——I 型胃神经内分泌肿瘤的推荐随访方案是什么？

An endoscopic follow-up every 12 months is recommended after complete endoscopic resection of a type I gNET. Whether this interval can be increased after prolonged periods without evidence of relapse or whether the interval should be shortened in case of incomplete (R1) resection or risk factors associated with progression (grading G2, size >20 mm) should be carefully discussed with the patient. In general, no cross-sectional imaging is required for follow-up. In addition, repeated evaluation of biomarkers CgA and gastrin should not be performed as these markers are elevated due to the underlying CAG and are not an indicator of relapse or progression. Follow-up OGD is recommended in a type I gNET that does not require resection, although the optimal interval for planning endoscopy is not established (1–2 years can be considered, following first follow-up 6–12 months after initial diagnosis). Apart from risk of tumour progression or relapse after resection in NET lesions, patients with CAG carry a relevant risk of developing gastric adenocarcinoma. Although only one international guideline recommends a regular endoscopic follow-up for patients with CAG, the yearly rate of adenocarcinoma detection is up to 1% in patients with type I gNETs.¹⁴ Whether this reflects an increased risk of adenocarcinoma development in patients with a previous diagnosis of gNET is not clear. For CAG patients without a gNET, intestinal metaplasia and H. pylori infection have been linked to an increased risk of adenocarcinoma progression. Given the high risk of adenocarcinoma detection upon follow-up of gNETs, this supports the recommendation to perform an endoscopic follow-up every 12–24 months in patients with previously diagnosed gNETs (specific timing may be proposed according to tumour size and histological features), whereas a 3-year interval is recommended in CAG patients without gNETs.²²
对于 I 型胃神经内分泌肿瘤（gNET）完全内镜切除后，建议每 12 个月进行一次内镜随访。若长期无复发证据，是否可延长随访间隔；或在不完全（R1）切除或存在进展风险因素（分级 G2、肿瘤大小>20 mm）时缩短间隔，应与患者充分讨论。通常随访无需进行横断面影像学检查。此外，不应重复检测生物标志物 CgA 和胃泌素，因这些指标会因基础慢性萎缩性胃炎（CAG）升高，不能提示复发或进展。对于无需切除的 I 型 gNET，建议进行随访性食管胃十二指肠镜检查（OGD），但最佳内镜间隔尚未确立（初次诊断后 6-12 个月首次随访，之后可考虑 1-2 年）。除 NET 病灶切除后的肿瘤进展或复发风险外，CAG 患者还面临发生胃腺癌的显著风险。 尽管仅有一项国际指南建议对慢性萎缩性胃炎（CAG）患者进行定期内镜随访，但在 I 型胃神经内分泌肿瘤（gNETs）患者中，每年腺癌检出率高达 1%。 ¹⁴ 这是否反映既往诊断为 gNET 的患者发生腺癌的风险增加尚不明确。对于未患 gNET 的 CAG 患者，肠上皮化生和幽门螺杆菌感染与腺癌进展风险增加相关。鉴于 gNETs 随访中发现腺癌的高风险，这支持对既往诊断为 gNETs 的患者每 12-24 个月进行一次内镜随访的建议（具体时间可根据肿瘤大小和组织学特征提出），而未患 gNETs 的 CAG 患者则建议每 3 年随访一次。 ²²

1.3.5 Q4 – When is surgery recommended for type I gNETs?
1.3.5 问题 4——何时建议对 I 型胃神经内分泌肿瘤进行手术治疗？

Although type I gNETs are usually treated by an endoscopic approach, upfront surgical resection is recommended in all tumours that are greater than 20 mm in size or with suspected muscolaris propria invasion (either on axial imaging or EUS). In addition, surgery could be considered in tumours showing high risk features on biopsy (e.g., high grade 2 NET [cutoff not established], lymphovascular invasion). A limited resection with sampling of local lymph nodes is the preferred approach. Total gastrectomy with D2 lymphadenectomy should be discussed in patients with known lymph node metastases or might be proposed as a completion procedure after final histology has proven lymphatic spread, although there are no solid data supporting this option.
尽管 I 型胃神经内分泌肿瘤通常采用内镜治疗，但对于所有直径大于 20 毫米或疑似固有肌层侵犯（通过轴向成像或超声内镜判断）的肿瘤，推荐首选手术切除。此外，对于活检显示高风险特征（如高级别 G2 NET[临界值未确定]、淋巴血管侵犯）的肿瘤，也可考虑手术。局限性切除联合局部淋巴结取样是首选方案。对于已知淋巴结转移的患者，应讨论全胃切除联合 D2 淋巴结清扫术，或在最终病理证实淋巴扩散后作为补充手术方案提出，但目前尚无确凿数据支持该选择。

Endoscopic resection should not be attempted in the presence of invasion into the muscularis propria, suspected lymph node metastases or high-risk features of metastatic spread (e.g., high Ki-67, vascular invasion, size >20 mm). These patients should receive full staging including cross-sectional imaging as described previously and ⁶⁸Ga-SSA-PET-CT and upfront surgical management. The optimal cutoff value for Ki-67 is not determined, but Ki-67 values above 10% should trigger evaluation of surgical treatment. In the rare instance of a G3 type I gNET a primarily surgical approach is recommended due to high risk of metastases.²³ Regarding tumour size, no clear cutoff for surgical management can be defined. Only limited data exist on tumour size and risk of metastases in the subgroup of type I gNETs. Data from the SEER database and from a large series from Taiwan showed that a cutoff of 20 mm was associated with an increased risk of lymph node metastases.^{13, 24}
对于已侵犯固有肌层、疑似淋巴结转移或存在转移高风险特征（如 Ki-67 高表达、血管侵犯、肿瘤大小>20 mm）的患者，不应尝试内镜下切除。这些患者应接受包括前述横断面成像、 ⁶⁸ Ga-SSA-PET-CT 在内的全面分期检查，并优先考虑外科手术治疗。Ki-67 的最佳临界值尚未确定，但 Ki-67 值超过 10%时应评估手术治疗的必要性。对于罕见的 I 型 G3 级胃神经内分泌肿瘤（gNET），由于高转移风险，推荐首选外科手术方案。 ²³ 关于肿瘤大小，目前无法明确界定外科治疗的绝对临界值。在 I 型 gNET 亚组中，关于肿瘤大小与转移风险的数据有限。来自 SEER 数据库和台湾地区大型病例系列的数据表明，20 mm 的临界值与淋巴结转移风险增加相关。 ^{13, 24}

In case of incomplete endoscopic resection (R1) or risk features in final pathology after endoscopic resection, a surgical approach should be discussed with the patient to achieve R0 resection and/or resection of local lymph nodes to exclude metastatic spread. In cases of lesions not amenable to an endoscopic approach (e.g., location close to cardia) a primary surgical resection might be necessary.
在内镜切除不完全（R1）或内镜切除后最终病理显示存在风险特征的情况下，应与患者讨论手术方案，以实现 R0 切除和/或局部淋巴结切除以排除转移扩散。对于不适合内镜处理的病灶（如靠近贲门的位置），可能需要进行初次手术切除。

SSA treatment to reduce gastrin production, thus diminishing the stimulus for further tumour growth and progression, could be proposed in patients who could not undergo resection or for multiple larger tumours or frequent relapse. Antrectomy should no longer be routinely offered but might be an option in patients not tolerating SSA or declining continuous application.
对于无法接受切除术、多发较大肿瘤或频繁复发的患者，可考虑采用 SSA（生长抑素类似物）治疗以降低胃泌素分泌，从而减少肿瘤进一步生长和进展的刺激。胃窦切除术不再作为常规推荐，但对于无法耐受 SSA 治疗或拒绝持续用药的患者，仍可作为备选方案。

1.5.1 Background  1.5.1 背景

Type III gNETs have been traditionally regarded as highly aggressive tumours requiring extended surgical resections (i.e., partial or total gastrectomy) with lymphadenectomy. However, with the widespread use of high-definition endoscopy, type III gNETs of smaller size and lower grade are increasingly being detected. A recent systematic review of 147 type III gNENs in which tumour management was reported found that ~45% were G1, 35% were G2 and 20% were G3.²⁶ As a result, less invasive treatment options, ranging from endoscopic resection to surgical wedge resection, have been investigated in highly selected patients. The management of type III gNETs is summarised in Figure 2.
III 型胃神经内分泌肿瘤（gNETs）传统上被视为高度侵袭性肿瘤，需进行扩大手术切除（如部分或全胃切除术）联合淋巴结清扫。然而，随着高清内镜的广泛应用，体积较小、分级较低的 III 型 gNETs 检出率日益增高。一项针对 147 例 III 型胃神经内分泌肿瘤（gNENs）治疗策略的系统综述显示，约 45%为 G1 级，35%为 G2 级，20%为 G3 级。 ²⁶ 因此，在严格筛选的患者中，研究者已探索了从内镜切除到外科楔形切除等创伤性更小的治疗方案。III 型 gNETs 的治疗策略总结见图 2。

1.5.2 Q5 – Can endoscopic treatments be proposed to treat type III gNETs?
1.5.2 问题 5 – 内镜治疗能否用于治疗Ⅲ型 gNETs？

Type III gNETs should be carefully characterised using endoscopy, biopsy, cross sectional thoracoabdominal CT and liver MRI, and often functional imaging (⁶⁸Ga-SSA-PET-CT or FDG/PET-CT depending on tumour grade) and in most cases EUS. Evidence from several retrospective case series suggests that carefully selected patients who have no evidence of lymph node involvement can be safely and successfully managed by endoscopic resection. Endoscopic resection appears to be most appropriate in patients who have localised G1 tumours ≤10 mm in diameter, because tumours greater than this size are more likely to have lymph node metastases (even though these may not be detected by imaging techniques).^{26, 27} However endoscopic resection can also be considered in patients with slightly larger tumours (<15 mm) and low G2 (Ki-67 3–10%) histology, particularly if the risks of surgical resection are considered to be high.²⁶ Current evidence does not support the use of a particular endoscopic resection technique (EMR vs. ESD).²⁶ Some very small (<5 mm) tumours have also been unintentionally removed by avulsion biopsy with no evidence of subsequent tumour recurrence, but this approach is not generally recommended.²⁶ Patients with positive resection margins (R1) should be considered for additional endoscopic resection or surgical salvage if appropriate. Patients who have undergone endoscopic resection should have close follow up with endoscopy and imaging (thoracoabdominal CT and liver MRI) to detect local and distant tumour recurrence.
Ⅲ型 gNETs 需通过内镜、活检、胸腹部 CT 平扫及肝脏 MRI 仔细评估，常需功能性影像（根据肿瘤分级选择 ⁶⁸ 镓标记生长抑素类似物 PET-CT 或 FDG/PET-CT），多数情况下还需超声内镜（EUS）。多项回顾性病例系列研究证据表明，经严格筛选且无淋巴结转移证据的患者可安全成功地接受内镜切除治疗。内镜切除最适用于直径≤10 毫米的局限性 G1 级肿瘤，因为更大尺寸肿瘤更可能伴淋巴结转移（即便影像技术未能检出）。 ^{26, 27} 但对于稍大肿瘤（<15 毫米）及低增殖活性 G2 级（Ki-67 3-10%）患者，若手术风险较高，亦可考虑内镜切除。 ²⁶ 现有证据未支持特定内镜切除技术（EMR vs. ESD）的优选性。 ²⁶ 部分极小的肿瘤（<5 毫米）也曾因活检撕脱被无意中切除，且未见后续肿瘤复发迹象，但此方法通常不予推荐。 ²⁶ 对于切缘阳性（R1）的患者，应考虑追加内镜切除或酌情进行手术补救。接受内镜切除的患者需通过内镜和影像学检查（胸腹 CT 及肝脏 MRI）密切随访，以监测局部及远处肿瘤复发。

1.5.3 Q6 – When should limited or extended surgical treatments be proposed to treat type III gNETs?
1.5.3 问题 6——何时应对 III 型胃神经内分泌肿瘤提出局限性或扩大性手术治疗方案？

Several clinicopathological features should be carefully evaluated to identify the best surgical option (limited vs. extended resection) for patients with type III gNETs.
需综合评估多项临床病理特征，以确定 III 型胃神经内分泌肿瘤患者的最佳手术方案（局限性切除 vs.扩大切除）。

A limited gastric wedge resection without standard lymphadenectomy can be considered as treatment option in patients with localised, G1 type III gNETs with no evidence of lymph node involvement on preoperative imaging (including EUS).²⁶ In patients fulfilling the above-mentioned criteria, tumour size, depth of infiltration and presence of lymphovascular invasion should be further assessed to define the extent of surgical resection. Therefore, a wedge resection can be safely proposed to patients with G1 type III gNETs <20 mm, limited to the submucosal layer and with no evidence of lymphovascular invasion.^{26, 28} The role of wedge resections in patients with G2 type III gNETs remains debated, as tumour grade represents a powerful predictor of disease aggressiveness. Similarly, the tumour size cutoff for proposing limited surgical resection is not clearly defined. Type III gNETs <10 mm are usually managed with endoscopic resection, whereas a limited surgical resection might be considered as initial treatment option for lesions measuring >10 mm and <20 mm.²⁶ Recent experiences report favourable oncological outcomes after endoscopic resection/limited surgical resection, thus strengthening the notion that a conservative approach might be appropriate for highly selected patients with type III gNETs.^{27, 29} In the presence of positive margins (R1) after endoscopic resection, salvage surgical wedge resection represents a possible treatment strategy.
对于术前影像学检查（包括超声内镜）未发现淋巴结受累的局限性 G1 型 III 类胃神经内分泌肿瘤患者，可考虑采用不进行标准淋巴结清扫的局限性胃楔形切除术作为治疗选择。 ²⁶ 对于符合上述标准的患者，应进一步评估肿瘤大小、浸润深度及是否存在淋巴血管侵犯以确定手术切除范围。因此，对于肿瘤直径<20 毫米、局限于黏膜下层且无淋巴血管侵犯证据的 G1 型 III 类胃神经内分泌肿瘤患者，可安全推荐楔形切除术。 ^{26, 28} 楔形切除术在 G2 型 III 类胃神经内分泌肿瘤患者中的应用仍存争议，因为肿瘤分级是疾病侵袭性的重要预测指标。同样，推荐局限性手术切除的肿瘤大小临界值尚未明确定义。直径<10 毫米的 III 类胃神经内分泌肿瘤通常采用内镜切除处理，而对于直径>10 毫米且<20 毫米的病灶，可考虑将局限性手术切除作为初始治疗选择。 ²⁶ 近期经验报告显示，内镜切除/局限性手术切除后肿瘤学结果良好，从而强化了对于经过严格筛选的 III 型胃神经内分泌肿瘤患者，保守治疗可能是合适选择的观点。 ^{27, 29} 在内镜切除后切缘阳性（R1）的情况下，补救性手术楔形切除是一种可能的治疗策略。

Radical surgical resection, either total or subtotal gastrectomy, with lymphadenectomy represents the procedure of choice for all the patients diagnosed with type III gNETs showing at least one of the following features: (1) nodal or distant metastases on preoperative imaging, (2) Ki-67 proliferative index defining a G3 tumour (Ki-67 >20%), and (3) tumour size >20 mm. Radical surgery is recommended as second-line treatment when final histology reveals one or more of the following findings: presence of nodal metastases, higher tumour grade compared with original biopsy, lymphovascular invasion or lack of complete tumour clearance (R1), in patients initially managed with a limited wedge resection.²⁶
对于诊断为 III 型胃神经内分泌肿瘤且至少符合以下任一特征的患者，根治性手术切除（全胃或次全胃切除术联合淋巴结清扫）是首选治疗方案：(1)术前影像学显示淋巴结或远处转移，(2)Ki-67 增殖指数界定为 G3 肿瘤（Ki-67 >20%），(3)肿瘤大小>20 mm。对于初始接受局限性楔形切除术的患者，若最终病理学检查发现以下一项或多项结果：存在淋巴结转移、肿瘤分级较原活检升高、淋巴血管侵犯或肿瘤未完全切除（R1），则推荐将根治性手术作为二线治疗。

1.5.4 Q7 – What scheduled follow-up is recommended after endoscopic/surgical resection of type III gNETs?
1.5.4 问题 7——III 型胃神经内分泌肿瘤内镜/手术切除后建议采用何种规范化随访方案？

Follow-up of patients who underwent surgical resection of type III gNETs is based on contrast-enhanced cross-sectional imaging (thoracoabdominal CT/liver-MRI with sometimes OGD and/or EUS or functional imaging [⁶⁸Ga-SSA-PET-CT and/or FDG-PET/CT depending on the tumour grade]). The timing of follow-up has never been clearly defined. When a total gastrectomy with lymphadenectomy is performed, the follow-up schedule adopted for gastric adenocarcinoma should be applied.²⁶ On the other hand, patients managed conservatively (endoscopic or surgical local excision) should undergo OGD after about 3 months to check the resection site and if this shows no macroscopic residual tumour, they should have regular follow-up with cross-sectional imaging and endoscopy/EUS. The frequency and choice of test will be influenced by the final tumour size and grade as well as patient fitness and in most cases, it will be possible to reduce the frequency of follow up as time progresses after resection. ⁶⁸Ga-SSA-PET-CT (or FDG-PET/CT depending on the tumour grade) and biopsies should be performed in the presence of a suspected disease relapse, but they are not routinely part of the follow-up programme.³⁰
对接受手术切除的 III 型胃神经内分泌肿瘤（gNETs）患者的随访基于增强横断面成像（胸腹部 CT/肝脏 MRI，有时结合 OGD 和/或 EUS 或功能成像[根据肿瘤分级选用 ⁶⁸ Ga-SSA-PET-CT 和/或 FDG-PET/CT]）。随访时间从未明确定义。若实施全胃切除加淋巴结清扫术，应采用胃癌的随访方案。 ²⁶ 另一方面，保守治疗（内镜或手术局部切除）的患者应在约 3 个月后接受 OGD 检查切除部位，若无宏观残留肿瘤，则定期进行横断面成像和内镜/EUS 随访。检查频率和选择取决于最终肿瘤大小、分级及患者身体状况，多数情况下可随术后时间推移减少随访频次。 ⁶⁸ Ga-SSA-PET-CT（或根据肿瘤分级选用 FDG-PET/CT）和活检应在疑似复发时进行，但非常规随访项目。 ³⁰

2.2.1 Pathology  2.2.1 病理学

Pathological assessment and grading of dNETs is like that of NETs of all other sites. Adequate sampling is needed and the deep mucosal location of these lesions, usually suspected by the endoscopist due to the smooth surface as opposed to the more common adenomas and carcinomas of the duodenum, warrants an attempt to perform deep biopsies. Morphologically, the main differential diagnosis in the duodenum is the CoGNET, previously classified as a ganglioneuroma or paraganglioma, but now shown to be more similar to NETs.^{36, 37}
十二指肠神经内分泌肿瘤（dNETs）的病理评估与分级与其他部位 NETs 相同。需进行充分取材，由于这类病变通常位于黏膜深层（内镜医师可通过其表面光滑特征与更常见的十二指肠腺瘤/癌进行鉴别），建议尝试深部活检。形态学上，十二指肠区主要需与节细胞性神经内分泌肿瘤（CoGNET）鉴别，该肿瘤曾归类为神经节瘤或副神经节瘤，现证实更接近 NETs。 ^{36, 37}

In cases with a high proliferation rate (i.e., mitotic count >20 in 2 mm² and/or Ki-67 index >20%), NEC and adenocarcinoma with neuroendocrine features should be ruled out (see guidance paper for digestive neuroendocrine carcinomas³⁵).
若增殖活性较高（即 2mm²内核分裂象>20 个 ² 和/或 Ki-67 指数>20%），需排除神经内分泌癌（NEC）及伴神经内分泌特征的腺癌（参见消化系统神经内分泌癌指南 ³⁵ ）。

2.2.2 Endoscopy  2.2.2 内镜检查

Most dNETs are in the first or second part of the duodenum.³⁸ Symptoms range from mostly none at incidental discovery to symptoms attributed to gastric outlet obstruction (in rare cases of large lesions) or anaemia. Patients with dNETs close to the ampulla and CoGNET which occur in the periampullary region can present with gastrointestinal bleeding, abdominal pain, anaemia or jaundice.^{38, 39} Periampullary dNETs are not infrequent in patients with neurofibromatosis type 1 (NF1), also known as Von Recklinghausen's disease including somatostatinomas.^{40, 41}
大多数十二指肠神经内分泌肿瘤（dNETs）位于十二指肠的第一或第二部分。 ³⁸ 症状从偶然发现的无症状到罕见大病灶引起的胃出口梗阻症状或贫血不等。靠近壶腹及壶腹周围发生的 CoGNET 患者可表现为消化道出血、腹痛、贫血或黄疸。 ^{38, 39} 壶腹周围 dNETs 在 1 型神经纤维瘤病（NF1，又称 Von Recklinghausen 病）患者中并不罕见，包括生长抑素瘤。 ^{40, 41}

At OGD, dNETs have similar features to gNETs but are typically single small sessile, erythematous or pale lesions in the duodenal cap, bulb or D1 to D2.⁶ In recent years, increasingly smaller lesions have been detected, owing to the better mucosal visualisation with modern endoscopic tools. The reported mean size of dNETs varies between 7 mm and 1.5 cm.⁴² Up to 10% of dNETs can occur as multiple tumours, which should prompt a diagnosis of MEN-I. It has been reported that approximately a quarter to a third of patients with dNETs and ZES, have undiagnosed MEN-I.^{43, 44}
胃镜检查中，dNETs 与胃神经内分泌肿瘤（gNETs）特征相似，但通常表现为十二指肠球部、球帽或 D1 至 D2 段的单发小型无蒂、红斑或苍白病灶。 ⁶ 近年来，随着现代内镜工具对黏膜可视化能力的提升，检测到的病灶越来越小。文献报道 dNETs 的平均大小在 7 毫米至 1.5 厘米之间。 ⁴² 多达 10%的 dNETs 可表现为多发性肿瘤，这种情况应提示多发性内分泌腺瘤病 I 型（MEN-I）的诊断。据报道，约四分之一至三分之一的 dNETs 合并胃泌素瘤（ZES）患者存在未确诊的 MEN-I。 ^{43, 44}

Endoscopic characterisation of dNETs has been described by Borbath and colleagues.⁶ Diagnosis is usually confirmed at endoscopic forceps biopsy (care as very small lesions may be almost completely excised at biopsy sampling and making future identification hard in case of R1 resection). Tumours of 1 cm or above should have accurate local staging by EUS as regional lymph node metastases occur in up to 40%–60% of cases, especially in duodenal gastrinomas.^{43, 45, 46}
Borbath 及其同事描述了 dNETs 的内镜特征。 ⁶ 诊断通常通过内镜钳活检确认（需注意极小的病灶可能在活检取样时几乎被完全切除，导致 R1 切除后难以再次识别）。对于 1 厘米或更大的肿瘤，应通过超声内镜（EUS）进行精确的局部分期，因为区域淋巴结转移发生率高达 40%-60%，尤其是十二指肠胃泌素瘤。 ^{43, 45, 46}

2.2.3 Imaging and laboratory examinations
2.2.3 影像学与实验室检查

Duodenal NETs are generally diagnosed by endoscopy and endoscopic ultrasound. Computed tomography after ingestion of water to fill stomach and duodenum, is recommended to help preoperative exploration. CT-enterography, MRI, ⁶⁸Ga-SSA-PET-CT and ¹⁸F-DOPA-PET/CT are not sufficiently accurate for localisation of well-differentiated gastroduodenal primary tumours,^47-49 but these examinations are used to help staging of regional and distant metastases.⁵⁰
十二指肠神经内分泌肿瘤（NETs）通常通过内镜和超声内镜诊断。建议术前采用饮水充盈胃十二指肠后的计算机断层扫描（CT）辅助探查。CT 小肠造影、磁共振成像（MRI）、 ⁶⁸ 镓标记生长抑素类似物 PET-CT 及 ¹⁸ 氟代多巴 PET/CT 对于高分化胃十二指肠原发肿瘤的定位准确性不足， ^47-49 但这些检查可用于评估区域性和远处转移的分期。 ⁵⁰

Laboratory tests are required for patients newly diagnosed with dNET. It is especially important to recognise potential gastrinoma (serum gastrin, chromogranin and occasionally secretin testing). Other laboratory tests, if symptoms are suggestive, may include somatostatin, and very rarely growth hormone releasing factor and cortisol with urinary 5-hydoxyindoleacetic acid.³⁸ If a diagnosis of genetic syndrome is suspected (e.g., MEN-I or NF1) specific laboratory tests and germline testing is advised^{38, 51} Distinguishing between dNET and neurofibromas may be difficult in some cases especially on axial imaging.
新确诊的十二指肠 NET 患者需进行实验室检测。尤其需识别潜在的胃泌素瘤（血清胃泌素、嗜铬粒蛋白及必要时胰泌素激发试验）。若存在相关症状，其他实验室检测可能包括生长抑素测定，极少数情况下需检测生长激素释放因子、皮质醇及尿 5-羟基吲哚乙酸。 ³⁸ 若怀疑遗传综合征（如多发性内分泌腺瘤病 1 型或神经纤维瘤病 1 型），建议进行特异性实验室检测和胚系基因检测。 ^{38, 51} 在某些病例中，尤其是轴位影像学检查时，区分十二指肠 NET 与神经纤维瘤可能存在困难。

2.3.1 Q8 – When is endoscopic resection indicated?
2.3.1 Q8 – 何时需要内镜下切除？

Very small non-functioning lesions 5 mm or less in D1 are commonly removed after lifting injection and snare (or by diagnostic biopsy) before any histological diagnosis is made and these lesions do not usually recur or metastasise.
直径 5 毫米或更小的 D1 区非功能性微小病灶通常在未进行组织学诊断前，通过注射抬举后圈套切除（或诊断性活检）即可清除，这些病灶通常不会复发或转移。

For younger and fitter patients who have lesions of 5–10 mm (and up to 15 mm in some centres), an opinion should be sought from therapeutic endoscopy about resection and the associated risks. EUS should be performed as well as axial and functional imaging. Most of these lesions are grade 1 and not invading the muscle layer. In these circumstances an endoscopic resection using EMR/Cap EMR/ESD is reasonable, but risks of perforation are significant, especially in D2. If endoscopic therapy is high risk (published rates of perforation 15%–25%) or unlikely to be curative, LR (i.e., duodenotomy with tumour excision or enucleation) or pancreatoduodenectomy (PD) (if considered essential by the surgeon) should be considered. Endoscopic therapy combined with laparoscopy for rescue in the event of perforation has been performed which may represent a novel and valuable alternative to ESD, able to guarantee a high R0 rate and a low risk of intraoperative duodenal perforation.⁵³
对于病灶大小为 5-10 毫米（部分医疗中心放宽至 15 毫米）且身体状况较好的年轻患者，应咨询治疗性内镜专家关于切除手术及相关风险的评估意见。需同时进行超声内镜（EUS）检查以及轴位和功能成像检查。此类病灶多为 1 级且未侵犯肌层。在此情况下，采用 EMR/帽式 EMR/ESD 进行内镜下切除是合理的，但穿孔风险较高（尤其在十二指肠第二段）。若内镜治疗存在高风险（文献报道穿孔率为 15%-25%）或难以达到根治效果，则应考虑局部切除（即十二指肠切开肿瘤切除术或剜除术）或胰十二指肠切除术（PD）（外科医生认为必要时）。已有研究采用内镜治疗联合腹腔镜穿孔补救方案，这可能成为 ESD 的新型替代方案，既能保证较高的 R0 切除率，又可降低术中十二指肠穿孔风险。 ⁵³

Duodenal NETs can be multiple, particularly if they are gastrinomas, and an association with MEN-I exists. Somatostatinomas can also occur in the duodenum, often close to the ampulla, but generally are without a clinical syndrome. There is an association with NF1 (functioning tumours are being addressed in a separate guidance paper on pancreatic functional NETs).
十二指肠 NETs 可为多发性，尤其当其为胃泌素瘤时，且与 MEN-I 存在关联。生长抑素瘤也可发生于十二指肠，通常靠近壶腹部，但通常不伴随临床综合征。该类肿瘤与 NF1 相关（功能性肿瘤的处理详见另一份关于胰腺功能性 NETs 的指南文件）。

2.3.2 Q9 – Can a “watch and wait” strategy be proposed in dNETs?
2.3.2 Q9 – 对于 dNETs，是否可以采取“观察等待”策略？

A watch and wait policy could be considered for very small (<5 mm) dNETs that cannot easily be resected, are non-functional, G1 and not invading the muscularis propria. This policy has usually been applied in patients unfit for endoscopic resection or surgery, since a more definitive therapy would be recommended in most cases, although a watch and wait policy has more recently been recommended as a result from some small series with limited follow-up. It is not clear if watch and wait is a suitable option in a patient who is fit for resection.^{54, 55}
对于无法轻易切除、无功能、分级为 G1 且未侵犯固有肌层的极小型（<5 毫米）十二指肠神经内分泌肿瘤（dNETs），可考虑采取观察等待策略。该策略通常适用于不适合内镜切除或手术的患者，因为在大多数情况下会推荐更明确的治疗方案，不过近期一些随访有限的小型研究也建议采用观察等待策略。目前尚不清楚观察等待是否适合适合切除的患者。 ^{54, 55}

2.3.3 Q10 – When is surgery indicated?
2.3.3 Q10 – 何时需要手术干预？

Data on nodal risk in very small non-functioning dNETs are scarce. Some series report nodal metastases as a common finding in patients with dNET (40%–60%), with tumour size being the most relevant risk factor for nodal involvement.⁵⁶ In cases with a tumour diameter >1 cm, nodal metastases were found in 13 out of 18 cases. Tumour location may be another important feature to be considered when defining surgical indication. Lesions arising in the ampullary/periampullary area differ from other dNETs due to a more aggressive biological behaviour in some studies⁵⁷ but not in others.⁵⁶ Functioning neoplasms harbour a higher metastatic potential.⁵⁸ Tumour invasion beyond the submucosa, tumour grade G2–G3 and lymphovascular invasion are factors affecting the risk of nodal metastases.
关于极小型无功能性十二指肠神经内分泌肿瘤（dNETs）淋巴结转移风险的数据较少。部分研究系列报告显示，dNET 患者淋巴结转移较为常见（40%–60%），其中肿瘤大小是淋巴结受累的最主要风险因素。 ⁵⁶ 在肿瘤直径>1 厘米的病例中，18 例中有 13 例发现淋巴结转移。肿瘤位置可能是决定手术适应症时需考虑的另一重要特征。起源于壶腹/壶腹周围区域的病变在某些研究中表现出更具侵袭性的生物学行为 ⁵⁷ ，但其他研究未得出相同结论 ⁵⁶ 。功能性肿瘤具有更高的转移潜能 ⁵⁸ 。肿瘤浸润超出黏膜下层、肿瘤分级 G2–G3 以及淋巴血管侵犯是影响淋巴结转移风险的因素。

Regarding the extent of resection, PD (pancreatoduodenectomy) and LR (local resection) represent the two main alternatives:
关于切除范围，PD（胰十二指肠切除术）和 LR（局部切除术）是两种主要选择：

In general, pancreatoduodenectomy with lymphadenectomy is the procedure of choice for ampullary/periampullary neoplasms, due to their particular aggressiveness. When nodal metastases are pre- or intraoperatively identified, PD with nodal dissection is indicated in order to obtain a proper oncological clearance. However, for lesions <20 mm, less aggressive approaches (i.e., local excision with lymphadenectomy or pancreas-sparing total duodenectomy) may be considered options. At least eight lymph nodes should be harvested to stage patients accurately.⁵⁹ However, radical surgery is associated with a high rate of short- and long-term complications. Therefore, when feasible from an oncological perspective, other surgical strategies should be considered.
一般而言，胰十二指肠切除术联合淋巴结清扫是壶腹部/壶腹周围肿瘤的首选术式，因其具有特殊的侵袭性。若术前或术中确认存在淋巴结转移，则需行胰十二指肠切除术联合淋巴结清扫以确保充分的肿瘤学切缘。但对于<20 毫米的病灶，可考虑创伤性较小的替代方案（如局部切除联合淋巴结清扫或保留胰腺的全十二指肠切除术）。为准确分期，至少应获取 8 枚淋巴结。 ⁵⁹ 然而，根治性手术伴随较高的短期及长期并发症发生率。因此，在肿瘤学允许的情况下，应考虑其他手术策略。

Local duodenal resection (i.e., duodenotomy with tumour excision or enucleation) represents a valuable and safe option, especially for patients with non-ampullary, non-functioning, dNET and without suspected nodal involvement at staging,⁵² (where endoscopic resection is not possible). The role of nodal dissection in this setting is unclear as recent experience does not support the clearance of occult nodal metastases due to lack of association with survival advantages.⁶⁰
局部十二指肠切除术（即十二指肠切开术联合肿瘤切除或剜除术）是一种有效且安全的选择，尤其适用于非壶腹、无功能、无淋巴结转移迹象的十二指肠神经内分泌肿瘤患者（当内镜下切除不可行时）。 ⁵² 在此情况下，淋巴结清扫的作用尚不明确，因为近期经验表明，由于缺乏生存获益关联，清除隐匿性淋巴结转移并无必要。 ⁶⁰

Recommendations  建议