INTRODUCTION —
The term "neuroendocrine tumor" (NET) refers to well-differentiated neuroendocrine neoplasms, and "neuroendocrine carcinoma" (NEC) refers to poorly differentiated neuroendocrine cancers. (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms".)
"Carcinoid syndrome" is the term applied to a constellation of symptoms mediated by various hormones and other vasoactive substances that are secreted by some NETs (table 1) [1]. Two of the most common manifestations are flushing and diarrhea (table 2). Most NETs are associated with carcinoid syndrome only when they have metastasized to the liver (table 3). (See "Clinical features of carcinoid syndrome".)
This topic will provide an overview of treatment of localized NETs, focusing on treatment of gastrointestinal primary sites. Related topics on the diagnosis and management of various neuroendocrine neoplasms are discussed separately.
●(See "Diagnosis of carcinoid syndrome and tumor localization".)
●(See "Treatment of the carcinoid syndrome".)
●(See "Diagnosis of carcinoid syndrome and tumor localization".)
●(See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms".)
●(See "Thymic neuroendocrine neoplasms".)
●(See "Well-differentiated high-grade (G3) gastroenteropancreatic neuroendocrine tumors".)
GENERAL PRINCIPLES —
In general, the basic principles of evaluation and management of patients with well-differentiated gastrointestinal NETs (table 4) include:
●Radiographic staging and tumor localization – Common imaging modalities include contrast-enhanced computed tomography (CT) or magnetic resonance imaging (MRI) scans as well as somatostatin receptor (SSTR)-based diagnostic imaging, such as gallium Ga-68 dotatate integrated positron emission tomography (PET)/CT.
Radiographic staging that includes both anatomic and SSTR-based imaging is obtained for all small-intestinal (jejunal/ileal NETs) and all grade 2 to 3 NETs. However, radiographic staging studies may be omitted in very low-risk cases; these include type 1 gastric NETs that are <1 cm and grade 1, and T1 rectal NETs that are <1 cm and grade 1. Nearly all patients with appendiceal NETs will have had a CT scan prior to appendectomy; further imaging can be omitted if tumors are <2 cm.
Mesenteric masses are usually indicative of a primary tumor located in the small intestine. Radiographic staging and tumor localization of NETs is addressed in detail separately. (See "Diagnosis of carcinoid syndrome and tumor localization".)
●Pathologic assessment of tumor differentiation and/or grade – (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms", section on 'Pathology and tumor classification'.)
●Removal of the tumor if all disease is surgically resectable, even if liver metastases are present. (See "Metastatic gastroenteropancreatic neuroendocrine tumors: Local options to control tumor growth and symptoms of hormone hypersecretion", section on 'Surgical resection'.)
●Control of carcinoid symptoms, if present. (See "Treatment of the carcinoid syndrome".)
STAGING EVALUATION
Staging studies — Multiphasic contrast-enhanced CT is used to evaluate all patients with well-differentiated gastrointestinal NETs, except for tumors with very low probability of spread, such as most type 1 gastric NETs, or small (<1 cm) superficial (T1) rectal NETs. (See "Diagnosis of carcinoid syndrome and tumor localization", section on 'Tumor localization and staging'.)
The utility of other staging studies and biochemical markers, such as chromogranin A (CgA) and urinary 5-hydroxyindoleacetic acid (5-HIAA), are discussed in detail separately. (See "Diagnosis of carcinoid syndrome and tumor localization" and "Clinical presentation, imaging and biomarker monitoring, and prognosis of metastatic well-differentiated gastroenteropancreatic neuroendocrine tumors" and "Diagnosis and staging of small bowel neoplasms", section on 'Neuroendocrine tumors' and "Overview of tumor biomarkers in gastroenteropancreatic neuroendocrine tumors".)
Staging system — Well-differentiated NETs of the gastrointestinal tract are staged by primary tumor site using the ninth version of the combined American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) tumor, node, metastasis (TNM) staging system.
●NETs of the appendix (table 5) [2]
●NETs of the jejunum and ileum (table 6) [3]
●NETs of the duodenum and ampulla of Vater (table 7) [4]
●NETs of the colon and rectum (table 8) [5].
●NETs of the stomach (table 9) [6]
TREATMENT —
The treatment of choice for a patient who has a localized well-differentiated NET is usually surgery, although endoscopic resection may be appropriate in certain circumstances. The extent of the surgical resection depends on the site of origin and size of the primary tumor. A detailed discussion of the clinical presentation of NETs arising in specific organs is available separately. (See "Clinical characteristics of well-differentiated neuroendocrine tumors arising in the gastrointestinal and genitourinary tracts".)
Appendix — The prognosis of appendiceal NETs is best predicted by tumor size. In many series, tumors less than 2 cm in diameter (found in approximately 95 percent of patients) have an extremely low likelihood of metastases at diagnosis. In contrast, up to 30 percent of larger tumors have already metastasized at diagnosis [7]. (See "Well-differentiated neuroendocrine tumors of the appendix", section on 'Clinical presentation'.)
Optimal surgical management for appendiceal NETs is subject to some debate. Because most are discovered incidentally in an appendectomy specimen done for other reasons, a decision must be made whether or not to return the patient to the operating room for a right colectomy. Unlike simple appendectomy, colectomy removes the draining lymph nodes of the appendix and any residual disease that might remain at the base of the appendix or in the mesoappendix.
Because of the association of tumor size with prognosis, tumors greater than 2 cm or those with mesoappendiceal invasion have been traditionally treated with right hemicolectomy [8]. On the other hand, whether or not a colectomy should be performed in some patients with smaller tumors is debatable. Although numerous studies have demonstrated a risk of local lymph node metastases associated with tumors measuring 1 to 2 cm in diameter, it is not clear whether lymph node involvement portends future distant metastases. As an example, a retrospective study of 278 patients with 1 to 2 cm appendiceal NETs demonstrated a very low rate of synchronous metastases (1 percent) and no cases of metachronous metastases, whether or not patients underwent completion right hemicolectomy [9]. There is general agreement that tumors less than 1 cm can usually be treated by simple appendectomy [8].
These issues are discussed in more detail separately. (See "Well-differentiated neuroendocrine tumors of the appendix", section on 'Treatment of localized disease'.)
Jejunum/ileum — NETs of the small intestine (jejunum/ileum, also known as midgut) are most commonly found in the ileum within 60 cm of the ileocecal valve. All small intestinal NETs have the potential to metastasize, irrespective of size and the large majority are associated with lymph node metastases [10].
Because of this, in our view, and that of others [8], patients with non-metastatic NET of the jejunal or ileum of any size should be treated with wide local excision that includes resection of the involved segment and small bowel mesentery. (See "Treatment of small bowel neoplasms", section on 'Neuroendocrine tumors'.)
Because multiple NETs are present in approximately 25 to 55 percent of cases, the remainder of the small bowel should be examined at the time of surgery [11,12]. Although this is a controversial area, resection of the primary tumor may be advised even in patients with known distant metastases in order to reduce the potential for bowel obstruction or bleeding, or to palliate abdominal pain related to the primary tumor; some reports note a beneficial impact on overall survival as well. (See "Metastatic gastroenteropancreatic neuroendocrine tumors: Local options to control tumor growth and symptoms of hormone hypersecretion", section on 'Management of the primary tumor in patients with metastatic disease'.)
There are limited data on the risk of metastatic recurrence after complete resection of small-intestinal NETs, including a lack of prospective studies with long-term follow-up in appropriately staged patients. In one observational study of patients with stage I to III resected small bowel NETs (of which 77 percent were lymph node positive), at median postoperative follow-up of 81 months, the rate of metastatic recurrence was 31 percent of patients [13]. The annual risk of recurrence was fairly constant for the first eight years postoperatively, with a subsequent decline in recurrence risk.
The prognosis depends upon disease stage [14]. (See 'Stage and site of origin' below.)
However, even among patients with distant metastasis, five-year overall survival rates range from 40 to 85 percent; 10-year survival rates of 40 to 60 percent are reported [14-18]. Ten-year disease-specific survival rates stratified according to the 2010 AJCC/UICC stage groupings from a series of 6792 patients with small intestine NETs diagnosed between 1988 and 2009 derived from the National Cancer Institute (NCI) Surveillance, Epidemiology, and End Results (SEER) registry were as follows [18]:
●Stage I – 95 percent (95% CI 93-97 percent)
●Stage IIA – 95 percent (95% CI 90-96 percent)
●Stage IIB – 77 percent (95% CI 71-83 percent)
●Stage IIIA – 68 percent (95% CI 58-77 percent)
●Stage IIIB – 77 percent (95% CI 74-80 percent)
●Stage IV – 42 percent (95% CI 38-46 percent)
Limited data are available that stratify prognosis according to the eighth edition American Joint Committee on Cancer (AJCC) staging system, which separated jejunoileal and duodenal/ampullary tumors and condensed stages I to IV disease (eliminating the A and B substages) [19].
Prognosis among surgically treated patients also depends on margin status [10,17]. (See 'Residual disease' below.)
Duodenum — Most localized duodenal NETs are detected incidentally in the duodenal bulb during upper endoscopy. The incidence of these tumors is rising due to increasing rates of endoscopy. Ampullary NETs tend to be more aggressive than duodenal NETs. Gastrinomas of the duodenum are discussed separately. (See 'Ampulla of Vater' below and "Management and prognosis of gastrinoma (Zollinger-Ellison syndrome)".)
Duodenal NETs are staged using endoscopic ultrasonographic (EUS), which influences management.
●Low-grade (G1) duodenal NETs <1 cm in diameter and confined to submucosal layer are often managed with endoscopic resection (using various techniques) since these tumors are at low risk of lymph node involvement or distant metastases [20]. However, the risk of perforation with endoscopic resection is higher in the duodenum than other parts of the gastrointestinal tract. Very small (<5 mm) grade 1 duodenal NETs can sometimes be resected endoscopically using simple snare polypectomy.
●The European Neuroendocrine Tumor Society (ENETS) recommends (with low degree of evidence) surgical resection for duodenal NETs that are >1 cm, invade the muscularis propria, are tumor grade G2 or G3, with lymphovascular invasion, and/or hormonally functioning [21]. The type of surgery may be controversial and depends on the exact location within the duodenum. While pancreaticoduodenectomy (ie, Whipple procedure) is often the definitive anatomic surgery, local duodenal resection (eg, duodenotomy with local resection) may be an option particularly if lymph node involvement is not suspected preoperatively.
Ampulla of Vater — There is a high likelihood of nodal metastases with ampullary NETs, even with tumor size <2 cm [22-25]. Pancreaticoduodenectomy has been suggested for resectable cases regardless of size due to relatively high risk of occult nodal metastases. However, these management strategy is based on limited case series, and treatment decisions may need to be individualized based on precise location of tumors, histologic grade, depth of invasion, and patient suitability for surgery.
Rectum — Most rectal NETs are small, localized, and mucosal or submucosal in depth [26,27]. As with rectal adenocarcinomas, transrectal endoscopic ultrasound (TEUS) is often useful for assessment of tumor size, depth of invasion, and lymph node involvement in rectal NETs [28,29]. (See "Clinical characteristics of well-differentiated neuroendocrine tumors arising in the gastrointestinal and genitourinary tracts", section on 'Rectum' and "Endoscopic ultrasound for evaluating patients with rectal cancer".)
While many small rectal NETs exhibit indolent behavior, certain risk factors predict for metastases. Traditionally, tumor size and depth of invasion (T stage) have been the main determinants of prognosis. Other prognostic risk factors include tumor grade, mitotic index, and lymphovascular invasion. As examples:
●One observational report [30] indicated that in tumors <2 cm, the rate of metastasis was 2 percent of patients if the tumor was confined to the submucosa, but as high as 48 percent if the tumor invaded the muscularis propria.
●Other reports [31,32] have identified additional poor prognostic factors even among patients with relatively small (<2 cm) and superficial tumors, including mitotic index and lymphovascular invasion (LVI) [33].
●An analysis of 199 patients identified risk factors for lymph node involvement, which were observed in 9 percent of patients in total. Tumor diameter >11.5 mm and Ki-67 >3.5 percent were optimal cutoff values predicting lymph node involvement. On multivariate analysis (which included Ki-67, tumor diameter, muscle invasion and vascular invasion), only tumor diameter >11.5 mm and vascular invasion were independently associated with lymph node metastases [34].
●In an observational study of 329 patients with rectal NETs <2 cm, risk factors for metastases included tumor size >1 cm, tumor grade G2 to G3, muscular invasion, and lymphovascular invasion [35].
Treatment options for localized rectal NETs include conventional endoscopic resection (standard polypectomy or endoscopic mucosal resection [EMR]), advanced endoscopic resection (endoscopic submucosal dissection, transanal endoscopic microsurgery [TEM], cap-assisted EMR), transanal surgical resection, or radical resection (low anterior resection [LAR], abdominoperineal resection [APR]). Decisions about technique should be based on tumor size and other risk factors for nodal metastases:
●Tumors smaller than 1 cm and confined to the mucosa or submucosa (T1) can generally be treated by standard endoscopic resection [26,36-40], particularly if they lack other risk factors (ie, mitotic rate >2 per 10 high-power fields [HPF], Ki-67 ≥3 percent or LVI). A greater likelihood of negative resection margins may be achievable with more advanced endoscopic techniques including endoscopic submucosal dissection [41,42] and cap-assisted EMR (in which the tumor is suctioned into a cap and then removed with a snare) [43]. However it is unclear whether negative margins are necessary given the excellent outcome of patients with low-grade T1 tumors, even with positive or indeterminant margins [40].
●Large tumors (>2 cm) or those that invade the muscularis propria (≥T2) are generally treated with surgical resection (LAR or APR) [44]. This approach is endorsed by consensus-based guidelines from the National Comprehensive Cancer Network (NCCN) and the ENETS [45,46]. Although some data suggest that local excision may be a viable alternative for some subgroups of patients with lower-risk rectal NETs >2 cm, the optimal way to select these patients is not established, and we do not endorse this approach unless patients refuse radical resection [47].
●The management of intermediate-sized tumors (1 to 2 cm confined to mucosa or submucosa) is somewhat controversial. Transanal resection or advanced endoscopic resection techniques (such as TEM) may be appropriate for tumors lacking risk factors, whereas formal resection may be more appropriate for tumors with risk factors such as elevated mitotic rate (or Ki-67 index), LVI, or size >1.5 cm [31,48].
For well-differentiated rectal NETs, prognosis is mainly dependent on tumor size and depth of invasion, as reflected by the T-stage classification [16]. Prognosis is also stratified according to disease stage [10,49-52]. As an example, in an observational study from the National Cancer Database of 258 well-differentiated rectal NETs that were stratified according to stage, five-year overall survival rates for stage I, II, III, and IV disease were 92, 88, 59, and 15 percent, respectively [53].
Colon — Colonic NETs tend to be more aggressive than those arising in the rectum. In one observational series from the SEER database, five-year survival rates were only 62 percent across all stages (compared with 88 percent for rectal NETs) [51]. In a dataset of 5457 colorectal NETs reported to the National Cancer Database between 1998 and 2002, compared with rectal primaries, a colonic primary site was associated with significantly poorer overall survival (hazard ratio [HR] for death 1.85, 95% CI 1.61-2.13) [53].
Well-differentiated colonic NETs (excluding ileocecum and rectum) are extremely rare. Although they appear to have the worst prognosis of any gastrointestinal tract well-differentiated NET [38,51,54,55], much of this literature is not contemporary and likely did not distinguish between both poorly differentiated colonic neuroendocrine carcinomas (NECs) and well-differentiated colonic NETs. [56,57]. (See "Clinical characteristics of well-differentiated neuroendocrine tumors arising in the gastrointestinal and genitourinary tracts", section on 'Colon'.)
Given that most tumors are >2 cm and/or invasive through the muscularis propria, patients with non-metastatic well-differentiated NETs of the colon should be managed with formal partial colectomy and regional lymphadenectomy, similar to treatment of colonic adenocarcinomas [38]. (See "Overview of the management of primary colon cancer", section on 'Surgical resection' and "Surgical resection of primary colon cancer".)
Prognosis is dependent on stage (table 8) [16]. Five-year overall survival for a series of 882 well-differentiated colonic NETs reported to the National Cancer Database between 1998 and 2002 and stratified according to AJCC stage were 86, 79, 65, and 27 percent for stage I, II, III, and IV disease, respectively [53].
Stomach — Gastric NETs have traditionally been subdivided into three categories, with differing biologic behavior and prognoses:
●Type 1 gastric NETs, which represent 70 to 80 percent of all gastric NETs, are associated with chronic atrophic gastritis. In this condition, serum gastrin rises in response to gastric achlorhydria. The elevated gastrin, in turn, stimulates neuroendocrine cell hyperplasia in the stomach and development of multifocal polypoid NETs. The clinical behavior of these tumors is usually indolent. Most are grade 1 tumors with stage I (superficially invasive and/or <1 cm in diameter) disease and no mortality with prolonged follow-up [58-60].
●Type 2 gastric NETs, which represent approximately 5 percent of gastric NETs, also occur as a result of elevated serum gastrin levels stimulating multifocal gastric NETs. The underlying cause of type 2 gastric NETs is a pancreatic or duodenal gastrinoma (Zollinger-Ellison syndrome). The clinical behavior is usually indolent.
●Type 3 (sporadic) gastric NETs occur in the absence of atrophic gastritis or the Zollinger-Ellison syndrome. They account for 20 percent of gastric NETs and are the most aggressive; local or hepatic metastases are present in up to 65 percent of patients who come to resection.
It is increasingly apparent that a fourth category of gastric NETs exists related to chronic use of proton-pump inhibitors (PPI). As in type 1 and type 2 disease, the mechanism is related to chronic hypergastrinemia, caused by suppression of gastric acid release [9,61,62].
Definitions and a more complete discussion of characteristics of the different types of gastric NETs are discussed separately. (See "Clinical characteristics of well-differentiated neuroendocrine tumors arising in the gastrointestinal and genitourinary tracts", section on 'Stomach'.)
Management depends on the type of gastric NET:
●Type 3 (sporadic) gastric NETs are generally treated by partial or total gastrectomy with local lymph node resection [63-65]. The risk of nodal metastases is dependent on tumor size and depth. Some have suggested that endoscopic resection alone may represent adequate therapy for tumors <1 cm invading the lamina propria or submucosa only [60,66], while others suggest wedge resection or endoscopic therapy alone only for those with a well-differentiated (grade 1) gastric NET no larger than 1.5 cm and without lymphovascular invasion [67]. However, this is not a standard approach, and we generally advocate gastrectomy/lymphadenectomy for all type 3 tumors, regardless of size and histologic differentiation.
For type 1 and 2 gastric NETs smaller than 1 to 2 cm, endoscopic resection represents adequate therapy [59,68,69]. Subsequent endoscopic surveillance is needed every 6 to 12 months since these patients continue to exhibit mucosal changes and hyperplasia of enterochromaffin-like cells (ECL) due to sustained hypergastrinemia. (See 'Post-treatment follow-up' below.)
Progression to a malignant phenotype or disease-related death is rare with small tumors [70]. Metastases occur in less than 10 percent of tumors ≤2 cm [71].
Antrectomy is a controversial option for type 1 gastric NETs if there are numerous progressive tumors. Antrectomy reduces hypergastrinemia by reducing the gastrin-producing cell mass in the antrum of the stomach; in most cases, this leads to tumor regression [68,72-75]. This approach was used in a series of 51 patients with type I gastric NETs, 10 of whom had antrectomy (eight in conjunction with endoscopic removal of the largest tumor) [68]. Seven of the eight with residual disease became endoscopically tumor-free, and one progressed and died of metastatic disease. In all, 9 of the 10 patients treated with antrectomy remained tumor free for an average of 65 months.
●More aggressive surgical therapy is rarely needed for type 1 gastric NETs unless there is extensive tumor involvement of the gastric wall (which increases the risk for a coexisting adenocarcinoma [76]), tumor size >2 cm (which increases the risk for metastases [71]), poorly differentiated histology, or emergent bleeding [77].
●The role of medical rather than surgical therapy (anti-gastrin maneuvers such as acidification by diet or dilute oral hydrochloric acid, or somatostatin analog therapy) for type 1 tumors is debated [63,74,78,79]. Gastrin levels may or may not decrease, and continued endoscopic surveillance is necessary.
●The optimal management of gastric NETs associated with chronic PPI use has not been well-established. Preliminary data indicate that these tumors are less prone to distant metastases than sporadic (type 3) gastric NETs and therefore less aggressive surgical approaches (such as endoscopic resection or surgical wedge resection) may be more appropriate in some cases than radical gastrectomy.
Lung — The preferred treatment for bronchial NETs is surgical resection. The clinical features, diagnosis, and treatment of bronchial NETs are discussed separately. (See "Lung neuroendocrine (carcinoid) tumors: Epidemiology, risk factors, classification, histology, diagnosis, and staging".)
POST-TREATMENT FOLLOW-UP —
There is limited evidence for the optimal follow-up after resection of a well-differentiated NET. For all resected well-differentiated small intestinal and colonic NETs and for rectal, gastric, and appendiceal NETs >2 cm, in our practice we focus on long-term (rather than frequent) surveillance given the fact that distant relapses can occur beyond five years after surgery, particularly with small bowel primary tumors. As an example, in one study of 129 midgut NETs, there was a fairly steady annual rate of recurrence for the first eight years after surgery, followed by a decline in risk of recurrence after eight years [13]. For patients with resected small intestinal NETs, we generally offer surveillance with triple-phase CT scans or MRI of the abdomen and pelvis every six months for the first year after surgery, then roughly annually for approximately 10 years after surgery [80]. For surveillance of NETs at other sites, we generally follow a similar strategy; however, more frequent surveillance can be considered for more aggressive tumors, such as grade 2, lymph node positive, gastric or rectal NETs. We do not offer post-treatment surveillance for low-grade appendiceal tumors <2 cm and superficial low-grade rectal tumors <1 cm in size given the exceptionally low risks of recurrence.
The role of nonhormonal biomarkers such as chromogranin A (CgA) in surveillance is controversial as the contribution to detection of recurrent disease is limited [13,81]. Furthermore, specificity is limited as a number of conditions and medications (especially proton pump inhibitors) can falsely elevate the CgA (table 10). Consensus-based guidelines have increasingly deemphasized the role of CgA in clinical care. (See "Overview of tumor biomarkers in gastroenteropancreatic neuroendocrine tumors", section on 'Role of nonhormonal tumor markers in clinical practice'.)
In general, 24-hour measurements of urinary 5-hydroxyindoleacetic acid (5-HIAA) are not obtained for postoperative surveillance of tumors that are unlikely to produce serotonin (ie, well-differentiated NETs of the hindgut [distal colon and rectum], or foregut [lungs, stomach, duodenum]). If serial assay of 24-hour urinary 5-HIAA is performed for the rare localized midgut tumor that was initially associated with a significant elevation in 5-HIAA, measurement requires strict adherence to dietary restrictions before and during urine collection. (See "Overview of tumor biomarkers in gastroenteropancreatic neuroendocrine tumors", section on 'Serotonin and 5-hydroxyindoleacetic acid (5-HIAA)'.)
Our approach to post-treatment surveillance is generally consistent with guidelines from the National Comprehensive Cancer Network (NCCN) [45], which are relatively non-specific after treatment of all resected small intestinal and colonic NETs and for rectal, gastric, and appendiceal NETs >2 cm:
●3 to 12 months postresection: Abdominal or abdominopelvic multiphasic CT or MRI scan for gastrointestinal NETs, chest CT with or without contrast for lung/thymus NETs. Assay of biochemical markers as clinically indicated.
●>1 year postresection: Every 12 to 24 months abdominal or abdominopelvic multiphasic CT or MRI scan for gastrointestinal NETs, chest CT with or without contrast for lung/thymus NETs. Assay of biochemical markers as clinically indicated.
●>10 years postresection: Continue surveillance as clinically indicated.
These surveillance strategies are particularly applicable for low and low-intermediate grade tumors (ie, Ki-67 <10 percent). More frequent surveillance and scan intervals may be appropriate for higher-grade NETs and particularly grade 3 NETs, particularly during the first two to three years after surgery.
For appendiceal tumors <2 cm without high-risk features and T1 low-grade rectal tumors <1 cm in size, we do not offer post-treatment surveillance given the exceptionally low risks of recurrence [82]. (See "Well-differentiated neuroendocrine tumors of the appendix", section on 'Posttreatment follow-up'.)
For T1 rectal tumors 1 to 2 cm in size, the only surveillance necessary is endorectal ultrasound or rectal MRI at 6 and 12 months [82].
For type I gastric NETs ≤2 cm, we offer esophagogastroduodenoscopy (EGD) every one to two years [82]. Routine imaging studies are not necessary and may be obtained only as clinically indicated.
Consensus-based follow-up guidelines are also available from a number of other groups including the Commonwealth Neuroendocrine Tumour Collaboration (CommNETs) [83], European Society for Medical Oncology (ESMO) [84], European Neuroendocrine Tumor Society (ENETS) [85,86] and North American Neuroendocrine Tumor Society (NANETS) [80,87].
PROGNOSIS —
Prognosis depends primarily on tumor stage, margin status, histologic grade/differentiation, and site of origin.
Stage and site of origin — Most data support the prognostic validity of tumor stage for well-differentiated NETs of the gastrointestinal tract [10,16,18,88,89]. However, at least some data suggest that there are no significant differences in outcomes between stage I and IIA midgut (jejunal and ileocecal) NETs (10-year overall survival 95 percent for both [18]), and heterogeneous outcomes in patients with stage IIIB (node positive) disease depending on whether disease was resected (five-year overall survival 95 percent) or unresectable (five-year overall survival 78 percent) [10]. (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms", section on 'Staging system'.)
In an observational series from Sweden of 135 surgically managed small intestine NETs, overall survival rates at fifteen years for patients with localized, regional, and distant metastatic disease was 42, 24, and 10 percent, respectively [17].
The primary tumor site also determines prognosis. As a general principle, well-differentiated NETs originating in the small intestine (midgut) are more prone to metastasize than are well-differentiated tumors originating elsewhere in the gastrointestinal tract (table 3). However, they also progress more slowly once they do metastasize; as a result, median survival durations are longer among patients with metastatic (stage IV) NETs of the small intestine compared with metastatic tumors originating in other sites [15]. Prognosis is particularly poor for colonic NETs [38].
The differing prognosis according to stage and site of origin was also supported in an observational study from the Surveillance, Epidemiology, and End Results (SEER) database that evaluated the impact of primary tumor site and stage (localized, locally advanced, metastatic) in 35,618 patients with well-differentiated gastrointestinal tract NETs [90].
Residual disease — Prognosis among surgically treated patients also depends on whether or not the resection was complete [10,17]. In an observational study from Sweden of 135 patients with surgically small intestinal NETs, an R0 (complete) resection was associated with improved 15-year disease-specific survival (78 versus 12 percent) and overall survival (33 versus 6 percent) relative to an R1 or R2 resection [17].
Tumor grade — For nearly all types of gastrointestinal NETs, indices of proliferative activity (ie, mitotic rate and Ki-67 index) correlate with recurrence risk as well as progression-free and overall survival. While the World Health Organization (WHO) classification system for neuroendocrine neoplasms of the digestive system establishes three categorical grades, the inverse relationship between proliferative activity and prognosis is fairly continuous (picture 1 and table 4) [91]. (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms", section on 'WHO classification system'.)
Issues related to grading of gastroenteropancreatic NETs are discussed in detail separately. (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms", section on 'Assessing tumor grade'.)
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Neuroendocrine neoplasms".)
SUMMARY AND RECOMMENDATIONS
●Staging imaging – Contrast-enhanced CT is used to evaluate all patients with well-differentiated localized gastrointestinal neuroendocrine tumors (NETs), except for tumors with very low probability of spread, such as most type 1 gastric NETs, or small (<1 cm) superficial (T1) rectal NETs. (See 'Staging studies' above.)
•Nearly all patients with appendiceal NETs will have had a CT scan prior to appendectomy; further imaging can be omitted if tumors are <2 cm. (See 'General principles' above.)
•Radiographic staging that includes both anatomic and somatostatin receptor (SSTR)-based imaging is obtained for all small-intestinal (jejunal/ileal NETs) and all grade 2 to 3 NETs. However, radiographic staging studies may be omitted in very low-risk cases such as type 1 gastric NETs thar are <1 cm and grade 1, and T1 rectal NETs that are <1 cm and grade 1. (See 'General principles' above.)
●Staging system – Well-differentiated NETs of the gastrointestinal tract are staged by primary tumor site using the combined American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) tumor, node, metastasis (TNM) staging system.
•NETs of the appendix (table 5)
•NETs of the jejunum and ileum (table 6)
•NETs of the duodenum and ampulla of Vater (table 7)
•NETs of the colon and rectum (table 8)
•NETs of the stomach (table 9)
●Management – For most patients with a localized well-differentiated gastrointestinal NET, we recommend resection (Grade 1B). The type and extent of surgical resection depends upon the site of origin and the size of the primary tumor. Resection of the primary tumor may be advised even in patients with known distant metastases in order to reduce the potential for bowel obstruction or bleeding, or to palliate abdominal pain related to the primary tumor. This is a controversial area, and decision making must be individualized. (See 'Staging evaluation' above.)
Management is based on the specific primary tumor site:
•Appendix – Simple appendectomy is usually sufficient for low-grade tumors <2 cm. A right hemicolectomy has been traditionally used to treat tumors >2 cm. It is unclear whether other potential risk factors, such as mesoappendiceal invasion, are indications for right hemicolectomy in patients with a tumor size of 1 to 2 cm. In general, appendectomy alone is appropriate for tumors <1 cm. (See 'Appendix' above and "Well-differentiated neuroendocrine tumors of the appendix".)
•Jejunum/ileum – Patients with jejunum/ileum NETs require resection of the involved segment and small bowel mesentery. (See 'Jejunum/ileum' above.)
•Duodenum – Low-grade (G1) duodenal NETs <1 cm in diameter and confined to submucosal layer are managed with endoscopic resection. Surgical resection is used to treat duodenal NETs that are >1 cm, invade the muscularis propria, tumor grade G2 or G3, have lymphovascular invasion, and/or hormonally functioning. Pancreaticoduodenectomy is often the definitive anatomic surgery. Local duodenal resection (eg, duodenotomy with local resection) may also be an option, particularly if lymph node involvement is not suspected preoperatively.
•Ampulla of Vater – Ampulla of Vater NETs may be more aggressive than other duodenal NETs. Pancreaticoduodenectomy has been advocated for resectable cases regardless of size, but treatment decisions need to be individualized. (See 'Ampulla of Vater' above.)
•Rectum – Rectal NETs that are smaller than 1 cm and confined to the mucosa or submucosa (T1) can be treated by local endoscopic excision. The management of tumors between 1 and 2 cm that are confined to the mucosa or submucosa is controversial, and treatment must be individualized based upon size and the presence of risk factors such as lymphovascular invasion and mitotic rate (or Ki-67 index). Tumors larger than 2 cm and those that invade into or beyond the muscularis propria or have regional lymph node metastases require low anterior resection (LAR) or abdominoperineal resection (APR). (See 'Rectum' above.)
•Colon – Most patients with non-metastatic colonic NETs should be managed with formal partial colectomy and regional lymphadenectomy. (See 'Colon' above.)
•Stomach – Local management for gastric NETs depends on the type (see 'Stomach' above):
-Type 3 (sporadic) gastric NETs are treated by partial or total gastrectomy with local lymph node resection.
-For type 1 and 2 gastric NETs smaller than 1 to 2 cm, endoscopic resection is the treatment of choice. More aggressive surgical therapy is rarely needed for type 1 gastric NETs unless there is extensive tumor involvement of the gastric wall (which increases the risk for a coexisting adenocarcinoma), tumor size >2 cm, or emergent bleeding.
-Gastric NETs associated with chronic proton-pump inhibitors (PPI) are generally treated with less aggressive surgical approaches (such as endoscopic resection or surgical wedge resection) rather than radical gastrectomy, since these tumors are less prone to distant metastases than type 3 gastric NETs.
●Post-treatment surveillance
•Given the long natural history and the propensity for small bowel NETs to metastasize, we offer surveillance with triple-phase CT scans or MRI of the abdomen or abdomen and pelvis 3 to 12 months after surgery, then roughly annually for approximately 10 years. A similar surveillance strategy can be pursued for rectal, gastric, and appendiceal NETs >2 cm, although more frequent surveillance should be considered for more aggressive tumors, such as grade 2 gastric or rectal NETs. (See 'Post-treatment follow-up' above.)
•For appendiceal tumors <2 cm without high-risk features and T1 low-grade rectal tumors <1 cm in size, we do not offer post-treatment surveillance. (See "Well-differentiated neuroendocrine tumors of the appendix", section on 'Posttreatment follow-up'.)
•For T1 rectal tumors 1 to 2 cm in size, the only surveillance necessary is endorectal ultrasound or rectal MRI at 6 and 12 months.
•For type 1 gastric NETs ≤2 cm, we offer esophagogastroduodenoscopy (EGD) roughly once a year. Routine imaging studies are not necessary and may be obtained only as clinically indicated.
●Prognosis – Prognosis depends primarily on tumor stage, margin status, histologic grade/differentiation, and site of origin. (See 'Prognosis' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges Stephen Goldfinger, MD, who contributed to an earlier version of this topic review.