Classification, clinical presentation, diagnosis, and staging of pancreatic neuroendocrine neoplasms

INTRODUCTION — 

Pancreatic neuroendocrine tumors (NETs) are rare neoplasms that arise in the endocrine tissues of the pancreas (picture 1). They can secrete various peptide hormones, including insulin, gastrin, glucagon, and vasoactive intestinal peptide, resulting in a myriad of clinical syndromes. Approximately 50 to 75 percent of pancreatic NETs are nonfunctioning (ie, unassociated with a hormonal syndrome).

The clinical presentation, diagnostic imaging studies, and staging of well-differentiated pancreatic NETs is presented here. Surgical management of sporadic pancreatic NETs, high-grade gastroenteropancreatic neuroendocrine neoplasms (NENs), functioning pancreatic NETs, and neuroendocrine neoplasms of unknown primary site are discussed separately.

●(See "Surgical resection of sporadic pancreatic neuroendocrine neoplasms".)

●(See "Well-differentiated high-grade (G3) gastroenteropancreatic neuroendocrine tumors".)

●(See "Poorly differentiated gastroenteropancreatic neuroendocrine carcinoma".)

●(See "Insulinoma".)

●(See "Glucagonoma and the glucagonoma syndrome".)

●(See "Zollinger-Ellison syndrome (gastrinoma): Clinical manifestations and diagnosis" and "Management and prognosis of gastrinoma (Zollinger-Ellison syndrome)".)

●(See "Clinical presentation, diagnosis, and management of VIPoma".)

●(See "Somatostatinoma: Clinical manifestations, diagnosis, and management".)

●(See "Neuroendocrine neoplasms of unknown primary site".)

CLASSIFICATION — 

The classification of pancreatic neuroendocrine neoplasms (NENs) has evolved considerably. The term "pancreatic neuroendocrine tumor" (NET) describes well-differentiated neoplasms regardless of histologic grade. The term "pancreatic neuroendocrine carcinoma" (NEC) denotes poorly differentiated histology (table 1). The term "islet cell tumor" (which denotes the presumed origination of pancreatic NENs in the islets of Langerhans) is no longer used.

Histologic grade — Measures of proliferative index (Ki-67 and mitotic index) are used to assign histologic grade for pancreatic NETs (table 1).

●Low-grade (grade 1) tumors are defined as having a mitotic index less than 2, and Ki-67 less than 3 percent.

●Intermediate-grade (grade 2) tumors are defined as having a mitotic index of 2 to 20 or Ki-67 of 3 to 20 percent (whichever highest).

●High-grade (grade 3) tumors are defined as having a mitotic index greater than 20 or Ki-67 greater 20 percent. All poorly differentiated NEC are high grade, typically with Ki-67 greater than 50 percent.

Full details on the pathology and classification of NENs arising from the digestive system are discussed separately. (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms".)

Functionality — Functionality also impacts classification. Pancreatic NETs that are functioning (ie, hormone secreting) are classified according to the predominant hormone they secrete and the resulting clinical syndrome. Nearly all functioning NENs are well-differentiated NETs (see "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms", section on 'Tumor functionality'):

●An insulin-producing pancreatic NET causing episodic hypoglycemia is classified as an insulinoma. (See "Hypoglycemia in adults without diabetes mellitus: Determining the etiology" and "Insulinoma".)

●A gastrin-producing tumor associated with Zollinger-Ellison syndrome is classified as a gastrinoma. (See "Zollinger-Ellison syndrome (gastrinoma): Clinical manifestations and diagnosis".)

●Other functioning pancreatic NETs that secrete glucagon (glucagonomas), somatostatin (somatostatinomas), or vasoactive intestinal polypeptide (VIPomas) are quite rare, with an estimated annual incidence of approximately 1 in 10 million. (See "Glucagonoma and the glucagonoma syndrome" and "Clinical presentation, diagnosis, and management of VIPoma" and "Somatostatinoma: Clinical manifestations, diagnosis, and management".)

Of note, immunohistochemical staining is not a defining criterion for tumor classification. For example, if a tumor stains for gastrin but does not produce symptoms of Zollinger-Ellison syndrome, it is not considered a gastrinoma.

The pathologic diagnosis of a functioning pancreatic NET should be the same as for a nonfunctioning NET, with the descriptive functional designation appended to the diagnosis when there is knowledge of a clinical syndrome. Although functionality may impact prognosis (eg, insulinomas are generally indolent tumors), the biologic behavior of most functioning pancreatic NETs is defined by the grade and stage of the tumor, as it is with nonfunctioning tumors.

EPIDEMIOLOGY — 

Pancreatic neuroendocrine tumors (NETs) are overall rare. The incidence of these tumors is ≤1 case per 100,000 individuals per year, and they account for 1 to 2 percent of all pancreatic tumors [1-3]. Pancreatic NETs represent less than 3 percent of primary pancreatic neoplasms [4]. Incidence rates have been increasing in the United States and elsewhere in the world, but it is possible that this is related to increased detection of asymptomatic disease on cross-sectional imaging and endoscopy done for other reasons [2,3,5].

Although they may manifest at any age, they most often occur in the fourth to sixth decades of life.

HEREDITARY SYNDROMES — 

Germline testing should be performed in all patients with a family history suggestive of a hereditary syndrome associated with pancreatic neuroendocrine tumors (NETs) (table 2). While not mandatory, germline testing is also an option in those with seemingly sporadic tumors, particularly in younger patients less than 50 years old, regardless of family history.

Most pancreatic NETs are sporadic, but they can be associated with hereditary endocrinopathies, including multiple endocrine neoplasia type I (MEN1), von Hippel-Lindau (VHL) disease, neurofibromatosis type I (NF1), and tuberous sclerosis (table 2). (See "Multiple endocrine neoplasia type 1: Clinical manifestations and diagnosis" and "Clinical presentation and diagnosis of von Hippel-Lindau disease", section on 'Pancreatic lesions' and "Neurofibromatosis type 1 (NF1): Pathogenesis, clinical features, and diagnosis" and "Tuberous sclerosis complex: Clinical features".)

Approximately 80 to 100 percent of patients with MEN1, up to 20 percent of patients with VHL, 10 percent of patients with NF1, and 1 percent of patients with tuberous sclerosis will develop a pancreatic NET within their lifetime [6]. Importantly, patients who develop a pancreatic NET in the context of an inherited syndrome must be evaluated separately regarding prognosis, as these tumors tend to be associated with a more indolent course than are sporadic tumors. Furthermore, specific treatments may also be available that are not useful for treatment of sporadic tumors (eg, the hypoxia-inducible factor 2 alpha inhibitor belzutifan for VHL-associated pancreatic NETs). (See "Surveillance and management of von Hippel-Lindau disease", section on 'Pancreatic solid lesions and pNETs'.)

Studies have also demonstrated an association between pancreatic NETs and other hereditary mutations in deoxyribonucleic acid (DNA) repair genes such as BRCA2, MUTYH and CHEK2 [7,8]. Approximately 15 to 20 percent of pancreatic NETs are associated with a hereditary syndrome, with a higher incidence in younger patients. While not mandatory, germline testing is an option for patients with seemingly sporadic pancreatic NET (particularly younger patients less than 50 years old), regardless of family history. (See "Genetic counseling: Family history interpretation and risk assessment".)

CLINICAL PRESENTATION — 

The clinical presentation of pancreatic neuroendocrine tumors (NETs) is changing. Most pancreatic NETs were previously described as functioning and were detected following diagnostic evaluation of a hormonal syndrome. In contrast, more contemporary clinical series describe most (between 50 and 85 percent) pancreatic NETs as nonfunctioning [9-17].

Functioning tumors — The clinical presentation of specific functional tumors is addressed in detail in separate UpToDate topics (table 3). In brief:

●Insulinoma – Insulinomas typically present with episodic hypoglycemia, which may cause confusion, visual change, unusual behavior, palpitations, diaphoresis, and tremulousness. Amnesia for hypoglycemia is common. (See "Insulinoma", section on 'Clinical presentation'.)

●Gastrinoma – Gastrinomas typically present with peptic ulcer disease; diarrhea can also be a prominent feature. (See "Zollinger-Ellison syndrome (gastrinoma): Clinical manifestations and diagnosis", section on 'Clinical manifestations'.)

●Glucagonoma – The clinical syndrome classically associated with glucagonoma includes necrolytic migratory erythema, cheilitis, diabetes mellitus, anemia, weight loss, diarrhea, venous thrombosis, and neuropsychiatric symptoms. (See "Glucagonoma and the glucagonoma syndrome", section on 'Clinical features'.)

●VIPoma – The main clinical features of VIPoma syndrome are watery diarrhea, hypokalemia, and hypochlorhydria. (See "Clinical presentation, diagnosis, and management of VIPoma", section on 'Clinical features'.)

Tumors presenting with a specific hormonal syndrome can develop additional syndromes subsequently due to secretion of more than one hormone [18].

Nonfunctioning tumors — Although nonfunctioning pancreatic NETs do secrete a number of substances, such as chromogranins, neuron-specific enolase, pancreatic polypeptide, and ghrelin, they do not present clinically with a hormonal syndrome as compared with their functional counterparts. As a result, they often present later in the course of the disease with symptoms of local compression or metastatic disease [19-21]. However, mean tumor diameter has decreased in the last two decades, attributed to the increased use of cross-sectional imaging, and a larger percentage of tumors (40 and 50 percent of nonmetastatic pancreatic NETs in two series [19,22]) are detected incidentally in asymptomatic patients who undergo diagnostic evaluations for unrelated conditions.

Occasionally, patients with tumors that appear to be nonfunctional develop symptoms of hormone secretion later in their disease course [18].

When symptomatic, the most common presenting symptoms of a nonfunctioning pancreatic NET are abdominal pain (35 to 78 percent), weight loss (20 to 35 percent), and anorexia and nausea (45 percent) [23-26]. Less frequent signs include obstructive jaundice (17 to 50 percent), intra-abdominal hemorrhage (4 to 20 percent), or a palpable mass (7 to 40 percent). Symptoms may also be attributable to metastatic disease. In a variety of published reports, between 32 and 73 percent of cases are metastatic at diagnosis [3,11,27-29].

Metastatic disease — Uncommonly, patients present with metastatic NET and an unknown primary site. The most common site of metastatic disease involvement for pancreatic NET is the liver [30]. (See "Neuroendocrine neoplasms of unknown primary site", section on 'Evaluation and management'.)

DIAGNOSTIC AND STAGING EVALUATION

Diagnostic approach — The diagnostic approach to a patient with a suspected or biopsy-proven pancreatic neuroendocrine tumor (NET) depends on the clinical scenario. Often, the diagnostic evaluation and staging imaging evaluation may overlap.

●The diagnosis of pancreatic NET is typically confirmed on biopsy and histopathologic evaluation of the obtained tumor tissue, either from the pancreatic primary or a site of distant metastatic disease. (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms".)

●Patients with a biopsy-proven pancreatic NET require appropriate staging imaging studies to determine the extent of disease spread, which influences selection of therapy. Other patients may have biopsy-confirmed metastatic disease (most often in the liver; other less common sites include retroperitoneal lymph nodes and bone [31]), but their primary site of disease has not been identified.

•For patients with a biopsy-proven well-differentiated pancreatic NET, or those with biopsy confirmed metastatic disease whose primary site has not yet been identified, we perform a helical triple-phase computed tomography (CT) (arterial phase, portal venous phase, noncontrast) or a multiphasic gadolinium-enhanced magnetic resonance imaging (MRI) of the abdomen to stage the extent of disease. The value of chest imaging is questionable. (See 'Computed tomography' below and 'Magnetic resonance imaging' below.)

Most patients also undergo baseline imaging using somatostatin receptor (SSTR) positron emission tomography (PET)-CT imaging that utilize radiotracers include gallium Ga-68 dotatate, gallium Ga-68 dotatoc (where available), and copper Cu-64 dotatate [32]. SSTR PET imaging is used for full-body staging as well for the detection occult primary tumors. SSTR PET imaging may be omitted in patients with a very small (eg, less than 1 to 1.5 cm) grade 1 pancreatic NET, where the risk of metastatic disease is very low. (See 'Somatostatin receptor-based imaging studies' below.)

●For patients who present with hormonal syndromes suspicious for a pancreatic NET but who lack evidence of disease on cross-sectional imaging or SSTR PET imaging, we perform endoscopic ultrasound (EUS). EUS with fine needle aspiration can also be used to establish a pathologic diagnosis, particularly in patients with early-stage disease. (See 'Endoscopic ultrasonography' below.)

●Occasionally, patients present with a hormonal syndrome suggestive of a pancreatic NET (such as hypoglycemia with confirmed hyperinsulinemia on a monitored fast) but lack evidence of disease on cross-sectional imaging, SSTR PET imaging, or EUS. These patients may require highly specialized localizing evaluations such as arterial stimulation with venous sampling (ASVS) if expertise in this technique is available. However, this diagnostic approach is rarely necessary in modern practice, as the combined use of imaging and endoscopic studies for an occult hormonally functioning pancreatic NET (such as a subcentimeter insulinoma) has improved the sensitivity of preoperative detection to nearly 100 percent [33]. (See 'Arterial stimulation with venous sampling' below and 'Intraoperative localization techniques' below and "Insulinoma", section on 'Tumor localization'.)

Imaging studies

Computed tomography — Multiphasic, contrast-enhanced CT of the abdomen is a standard modality for evaluation of patients with pancreatic NETs.

Most NETs are highly vascular, and liver metastases may appear isodense with the liver on a noncontrasted study. Following the injection of intravenous contrast, pancreatic NETs often enhance with iodinated contrast during the early arterial phase (approximately 20 seconds after contrast injection), with washout during the portal venous imaging phase (approximately 70 seconds after contrast injection) [34]. In addition, arterial phase and portal venous phase sequences can be used to maximize the conspicuity of liver metastases compared with the surrounding normal liver parenchyma (image 1) [35].

CT scans are highly accurate for detecting primary pancreatic NETs; using modern multiphase imaging techniques, sensitivity is >80 percent [35-37]. Tumors as small as 4 mm have been visualized with CT scanning; however, sensitivity is decreased for tumors smaller than 2 cm in diameter compared with larger tumors [36]. Small tumors can often appear as rounded, enhancing vascular lesions (image 2); others may be hypodense or cystic (image 3).

Symptomatic but nonfunctioning tumors, VIPomas, and glucagonomas are usually large (>3 cm) at the time of diagnosis. The sensitivity of contrast-enhanced CT for these tumors approaches 100 percent, and it is considered the imaging study of choice [38,39].

Magnetic resonance imaging — On gadolinium-enhanced MRI of the abdomen, pancreatic NETs are typically characterized by low signal intensity on T1-weighted images (image 4) and by high signal intensity on T2-weighted images (image 5). The sensitivity of MRI to detect NETs has improved substantially with contemporary techniques, such as short tau inversion recovery (STIR) sequences [40-42]:

●In one study of 28 consecutive patients with a clinically suspected functioning pancreatic NET, MRI detected a pancreatic NET in 17 of 20 patients in whom it was done (sensitivity 85 percent) [40]. Specificity was 100 percent, and the positive and negative predictive values were 100 and 73 percent, respectively. In this study, the gold standard for diagnosis of a pancreatic NET was either resection (n = 19) or clinical follow-up for at least a year.

●In another study of 64 patients with metastatic NETs, multiphasic MRI detected more hepatic lesions than either contrast-enhanced CT or SSTR scintigraphy [43]. As a result of this greater sensitivity for liver metastases, some clinicians prefer MRI over CT for assessing the status of the liver [42]. This subject is discussed in detail separately. (See "Clinical presentation, imaging and biomarker monitoring, and prognosis of metastatic well-differentiated gastroenteropancreatic neuroendocrine tumors", section on 'Cross-sectional imaging (CT and MRI)'.)

As with CT scans, early arterial phase imaging following the injection of gadolinium contrast is critical for the detection of small hypervascular liver metastases. In a study of 37 patients with liver metastases from gastroenteropancreatic NETs, the most sensitive sequences for detection of liver metastases were hepatic arterial phase and fast spin-echo T2-weighed images (image 5) [37].

Endoscopic ultrasonography — EUS provides high-resolution imaging of the pancreas, and it can detect lesions as small as 2 to 3 mm in diameter (image 6). The combined use of imaging and endoscopic studies for an occult hormonally functioning pancreatic NET (such as a subcentimeter insulinoma) has improved the sensitivity of preoperative detection to nearly 100 percent [33]. (See "Insulinoma", section on 'Tumor localization'.)

Studies of EUS for detecting pancreatic NETs suggest high sensitivity for tumor detection [33,36,41,42,44-48]:

●In a systematic review of 17 cohort studies (612 patients), when it was performed, EUS identified pancreatic NETs in 97 percent of cases [48]. When the analysis was limited to cases where the pancreatic NET was not detected by CT with or without other modalities, a pancreatic NET was diagnosed solely by EUS in 28 percent of cases. Studies in which multidetector CT and/or spiral CT scans were known to have been performed before EUS noted a smaller incremental benefit for EUS.

●In another report, EUS was more sensitive than either CT or transabdominal ultrasonography for detection and localization of pancreatic NETs in patients with multiple endocrine neoplasia type 1 (MEN1) syndrome [47]. In this report, sensitivity for tumors less than 6 mm in diameter was poor, as determined by inspection of resected specimens by pathologists. However, others suggest acceptable reproducibility for EUS in detecting small pancreatic NETs in patients with MEN1 [49].

●EUS has also proven to be a useful tool for identifying pancreatic NETs such as gastrinomas that arise in the duodenal wall and have a high frequency of metastasis to the peripancreatic lymph nodes [50]. Duodenal gastrinomas are notoriously difficult to localize by CT.

Another benefit of EUS is that EUS-guided fine needle tissue sampling can often provide a nonoperative histologic diagnosis of pancreatic NET [51-53]. (See "Endoscopic ultrasound-guided fine needle aspiration in the gastrointestinal tract", section on 'Pancreatic lesions' and "Endoscopic ultrasound-guided fine needle biopsy in the gastrointestinal tract", section on 'Pancreatic lesions'.)

EUS is limited by the requirement of a highly skilled endoscopist and by its inability to consistently visualize the pancreatic tail. Nevertheless, this technique has proven useful for the localization of primary pancreatic NETs in patients without metastatic disease.

Somatostatin receptor-based imaging studies — Most well-differentiated NETs express high levels of somatostatin receptors (SSTRs) and can be imaged with radiolabeled somatostatin analogs. These scans allow for whole-body imaging. They also provide information on SSTR expression, which has important therapeutic implications regarding use of cold and radiolabeled somatostatin analogs.

SSTR PET-CT imaging is most frequently used due its superior sensitivity and image resolution. Several PET tracers are available, such as Ga-68 dotatate and Cu-64 dotatate. These tracers, in combination with integrated PET-CT, improve the detection and staging of NETs over prior methods of SSTR imaging such as OctreoScan [32,54-56]. SSTR PET imaging is useful for initial full-body staging as well as for detection of occult primary tumors. This subject is discussed in detail separately. (See "Clinical presentation, imaging and biomarker monitoring, and prognosis of metastatic well-differentiated gastroenteropancreatic neuroendocrine tumors", section on 'Somatostatin receptor-based imaging studies'.)

The affinity of SSTR-based imaging is highest for subtype 2 somatostatin receptors (SSTR2). Insulinomas, which are the most frequent type of functioning pancreatic NET, express relatively scant levels of SSTR2 and may be less likely to be detected with these scans [57-60].

Poorly differentiated neuroendocrine carcinomas (table 1) also express low SSTR levels and SSTR PET scans are not recommended for detection or staging of NEC. (See "Poorly differentiated gastroenteropancreatic neuroendocrine carcinoma".)

Arterial stimulation with venous sampling — For rare cases in which radiographically occult, hormonally functional tumors elude detection by conventional imaging modalities, invasive approaches may be employed to localize tumors to a particular region of the pancreas for treatment planning purposes (eg, tail, body/neck, head/uncinate). These include transhepatic portal venous sampling (THPVS) and ASVS [38,44]. These procedures have become relatively obsolete given the improved sensitivity of preoperative imaging, including SSTR PET/CT and EUS.

Intraoperative localization techniques — Intraoperative ultrasonography allows high-resolution examination of the pancreas. When combined with palpation of the organ, the sensitivity for tumor detection ranges from 83 to 100 percent [38,61,62]. Intraoperative transillumination has equivalent efficacy (sensitivity 83 percent) for the localization of duodenal wall gastrinomas [61].

Neither intraoperative ultrasonography nor illumination should replace preoperative imaging; they may be used as adjuncts to intraoperative palpation in patients who still have a suspected hormone-secreting pancreatic NET that cannot be identified or localized preoperatively.

Laboratory studies — A variety of peptide markers, including chromogranin A (CgA) and pancreatic polypeptide (PP), may be used in the follow-up of patients with pancreatic NETs. However, due to suboptimal sensitivity and specificity (particularly for CgA in the setting of proton pump inhibitor use), the value of routine tumor marker measurement has been questioned. Guidelines from the National Comprehensive Cancer Network (NCCN) consider assay of either CgA or PP for any pancreatic NET to represent a class 3 recommendation (there is major disagreement within the guidelines panel that the intervention is appropriate).

Functioning tumors — Specific hormones secreted by functional tumors (ie, insulin, proinsulin, glucagon, gastrin, vasoactive intestinal polypeptide) can be measured and correlated with hormonal symptoms. Hormone levels also correspond to changes in tumor burden and can therefore serve as specific tumor markers. (See "Insulinoma" and "Somatostatinoma: Clinical manifestations, diagnosis, and management" and "Clinical presentation, diagnosis, and management of VIPoma" and "Glucagonoma and the glucagonoma syndrome", section on 'Serum glucagon' and "Overview of tumor biomarkers in gastroenteropancreatic neuroendocrine tumors", section on 'Hormones associated with pancreatic NETs'.)

Nonfunctioning tumors — We do not routinely test nonspecific panels of hormone levels or nonhormonal tumor biomarkers such as CgA for patients with nonfunctioning tumors, which account for most pancreatic NETs. (See "Overview of tumor biomarkers in gastroenteropancreatic neuroendocrine tumors".)

STAGING SYSTEM — 

Pancreatic neuroendocrine tumors (NETs) are staged using the combined American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) staging system, which is based on the definitions proposed by the European Neuroendocrine Tumor Society (ENETS) [63]. These staging systems are highly prognostic for both relapse-free and overall survival [64-66].

Well-differentiated pancreatic NETs are staged using the ninth version of the AJCC/UICC tumor, node, metastasis (TNM) staging system (table 4) [67] that is separate from that used for exocrine pancreatic tumors [68]. Importantly, this staging system does not apply to high-grade, poorly differentiated neuroendocrine carcinomas; these tumors are staged according to the AJCC/UICC staging system for exocrine pancreatic cancer (table 5).

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

●Classification and nomenclature – The classification of pancreatic neuroendocrine neoplasms (NENs) has evolved considerably. (See 'Classification' above.)

•Pancreatic NET – The term "pancreatic neuroendocrine tumor" (NET) describes well-differentiated NENs regardless of histologic grade.

•Pancreatic NEC – The term "pancreatic neuroendocrine carcinoma" (NEC) is reserved for those cases with poorly differentiated histology, which is invariably associated with highly elevated Ki-67 and/or mitotic rate (table 1).

●Clinical presentation – Most pancreatic NETs were previously described as functioning and were detected following diagnostic evaluation of a hormonal syndrome. More contemporary clinical series describe most pancreatic NETs (50 to 85 percent) as nonfunctioning. (See 'Clinical presentation' above.)

●Hereditary syndromes – Germline testing should be performed in all patients with a family history suggestive of a hereditary syndrome associated with pancreatic NET (table 2). While not mandatory, germline testing is also an option in those with seemingly sporadic tumors, particularly in younger patients less than 50 years old, regardless of family history. (See 'Hereditary syndromes' above.)

●Diagnostic evaluation – The diagnostic approach to a patient with a suspected or biopsy-proven pancreatic NET is as follows (see 'Diagnostic approach' above):

•The diagnosis of pancreatic NET is typically confirmed on biopsy and histopathologic evaluation of the obtained tumor tissue, either from the pancreatic primary or a site of distant metastatic disease. (See "Pathology and classification of gastroenteropancreatic neuroendocrine neoplasms".)

•For patients with a biopsy-proven well-differentiated NET in the pancreas, or those with biopsy-confirmed metastatic disease whose primary site has not yet been identified, we perform a helical triple-phase CT (arterial phase, portal venous phase, noncontrast) or a multiphasic gadolinium-enhanced MRI of the abdomen to stage the extent of disease. In most cases, we also obtain baseline somatostatin receptor (SSTR) positron emission tomography (PET) imaging, which can be used for full-body staging as well as the detection of occult primary tumors. SSTR PET imaging may be omitted in patients with a very small (eg, less than 1 to 1.5 cm) grade 1 pancreatic NET at low risk of metastases. (See 'Imaging studies' above.)

•For patients who present with hormonal syndromes suspicious for a pancreatic NET but lack evidence of disease on cross-sectional imaging or SSTR PET imaging, we perform endoscopic ultrasound (EUS). EUS with fine needle tissue sampling can also be used to establish a pathologic diagnosis, particularly in patients with early-stage disease. (See 'Endoscopic ultrasonography' above.)

•In rare cases where there is strong clinical and biochemical evidence for an occult pancreatic NET that is undetectable on cross-sectional imaging, SSTR PET imaging, or EUS, we perform arterial stimulation with hepatic venous sampling (ASVS) if expertise in this technique is available, but this diagnostic approach is rarely necessary. (See 'Arterial stimulation with venous sampling' above.)

●Imaging studies

•CT or MRI abdomen – Multiphasic contrast-enhanced CT or gadolinium-enhanced MRI of the abdomen is highly sensitive for identification of primary pancreatic NETs as well as liver metastases. Early arterial phase imaging is particularly valuable for detection of hypervascular primary tumors and liver metastases. (See 'Computed tomography' above and 'Magnetic resonance imaging' above.)

•SSTR PET-CT – SSTR-based PET-CT imaging offers whole-body imaging and functional information regarding tumoral expression of SSTRs. Tracers include gallium Ga-68 dotatate and copper Cu-64 dotatate. The sensitivity of SSTR-based imaging may be lower for the detection of insulinomas and poorly differentiated or high-grade tumors. (See 'Somatostatin receptor-based imaging studies' above.)

•EUS – EUS is highly sensitive for detection of occult, subcentimeter pancreatic NETs and plays an important role in the evaluation of patients with functional NETs that are undetectable using conventional imaging techniques. Through endoscopic fine needle tissue sampling, EUS also offers the ability to obtain a nonoperative histopathologic diagnosis. (See 'Endoscopic ultrasonography' above.)

●Invasive localizing techniques – Among patients presenting with hormonal syndromes such as hypoglycemia or Zollinger-Ellison syndrome, the diagnostic modalities listed above (CT, MRI, somatostatin-receptor-based imaging, EUS) can detect nearly 100 percent of pancreatic NETs. As a result, invasive localizing techniques (such as arterial stimulation with venous sampling [ASVS]) are rarely necessary in modern practice. (See 'Arterial stimulation with venous sampling' above.)

●Staging – Well-differentiated pancreatic NETs are staged using the ninth version of the American Joint Committee on Cancer (AJCC)/Union for International Cancer Control (UICC) tumor, node, metastasis (TNM) staging system (table 4), which is separate from that used for exocrine pancreatic tumors. High-grade, poorly differentiated neuroendocrine carcinomas are staged according to the AJCC/UICC staging system for exocrine pancreatic cancer (table 5). (See 'Staging system' above.)

ACKNOWLEDGMENT — 

The editorial staff at UpToDate acknowledge Stephen Goldfinger, MD, who contributed to an earlier version of this topic review.