Islet Cell Tumors of the Pancreas




Islet cell tumors of the pancreas, also known as pancreatic neuroendocrine tumors, constitute less than 5% of pancreatic tumors, and 7% of all neuroendocrine tumors. Most are non-functional, and patients often present with metastatic disease. Functional tumors present with distinct clinical syndromes. Accurate staging is critical as surgery is both the cornerstone of treatment, and the only hope for cure. Medical management involves treating the manifestations of hormonal excess, and using somatastatin analogues when appropriate. Systemic chemotherapy, targeted molecular therapy, and peptide receptor radiotherapy may be used for refractory disease in lieu of or as an adjunct to surgery.


Key points








  • Islet cell tumors of the pancreas, also known as pancreatic neuroendocrine tumors (PNETs), are rare neoplasms that constitute fewer than 5% of pancreatic tumors.



  • Although PNETs may present with distinct clinical syndromes, most of these tumors are nonfunctional.



  • In a minority of cases, PNETs arise in the background of known genetic syndromes such as multiple endocrine neoplasia type I or von Hippel-Lindau syndrome.



  • Surgery remains the cornerstone of treatment, provided that 90% of the tumor burden can be resected.



  • For nonresectable well-differentiated lesions, somatostatin analogues, chemotherapy, targeted molecular therapy, and peptide receptor radiotherapy are appropriate options that lead to increased survival.






Introduction


Islet cell tumors of the pancreas, also known as pancreatic neuroendocrine tumors (PNETs), are a group of tumors that arise from the endocrine pancreas. Although PNETs may produce distinct clinical syndromes, most of these neoplasms are asymptomatic and present as an incidental finding. PNETs are a distinct group of neuroendocrine tumor (NET), with important biological differences from luminal (carcinoid) neoplasms. This article summarizes the important characteristics of PNETs with regard to epidemiology, pathology, diagnosis, and treatment, with a focus on new developments and their potential roles in the evolving management of this disease.




Introduction


Islet cell tumors of the pancreas, also known as pancreatic neuroendocrine tumors (PNETs), are a group of tumors that arise from the endocrine pancreas. Although PNETs may produce distinct clinical syndromes, most of these neoplasms are asymptomatic and present as an incidental finding. PNETs are a distinct group of neuroendocrine tumor (NET), with important biological differences from luminal (carcinoid) neoplasms. This article summarizes the important characteristics of PNETs with regard to epidemiology, pathology, diagnosis, and treatment, with a focus on new developments and their potential roles in the evolving management of this disease.




Epidemiology


PNETs are rare neoplasms that constitute fewer than 5% of pancreatic tumors, and only 7% of all NETs. Nevertheless, in both Europe and North America, the incidence of PNETs is rising. Data from the Surveillance, Epidemiology, and End-Results Program suggest an incidence of 0.43/100,000 person-years today, compared with 0.17 in the 1970s. This increase may be due to advances in diagnostic imaging and increased utilization of these technologies. Interestingly, autopsy studies have reported a widely varied incidence from 0.07% to as high as 10%, suggesting that many of these tumors are clinically silent and often go undiagnosed. Approximately 90% of PNETs diagnosed in the United States are nonfunctioning. PNETs can present at any age; however, the incidence peaks in the sixth and seventh decades. Not surprisingly, functional tumors present earlier than nonfunctional tumors (mean age 55 vs 59 years).




Syndromic association


Although 90% of PNETs occur sporadically, these tumors are also well-recognized features of 4 familial syndromes: multiple endocrine neoplasia type I (MEN1), von Hippel-Lindau syndrome (VHL), neurofibromatosis type 1 (NF1), and tuberous sclerosis complex (TSC). Each of these syndromes is inherited in an autosomal dominant pattern, and the causative genes are MEN1 , VHL , NF1 , and TSC1/2 , respectively. Whereas most patients (80%–100%) with MEN1 develop PNET, the frequency is much lower among the 3 other syndromes. Among patients with VHL, NF1, and TSC, 10% to 17%, 0% to 10% (almost all duodenal somatostatinomas), and fewer than 1%, respectively, develop PNETs. In general, PNETs that arise in the background of a familial syndrome tend to follow a more indolent course than sporadic tumors.


Multiple Endocrine Neoplasia Type I


MEN1 is characterized by PNETs in association with pituitary and parathyroid tumors. Almost all patients with MEN1 (>95%) will develop a PNET during their lifetime, although most of these will be nonfunctioning “micro-adenomas” (smaller than 0.5 cm) that are typically multifocal. Fewer than 15% of these nonfunctioning tumors will be large enough to be symptomatic. Functioning PNETs that are symptomatic occur in between 20% and 70% of patients with MEN1, with approximately 55% of these patients presenting with Zollinger-Ellison (ZE) syndrome due to an underlying gastrinoma, and 20% presenting with symptoms from an insulinoma. Other functioning PNETs, such as VIPoma, glucagonoma, and somatostatinoma occur in fewer than 3% of patients with MEN1. The management of patients with MEN1 with PNETs is of particular significance, as PNETs are the leading cause (40%) of disease-specific mortality among patients with MEN1. Furthermore, the mean age of death among patients with MEN1 with PNETS is 55 years, which is lower than that of both the general population and patients with non-MEN1 PNETs.


Von Hippel-Lindau Syndrome


Although pancreatic lesions are common in VHL, only 10% to 17% of patients with VHL develop PNETs. Furthermore, almost all VHL-associated PNETs are nonfunctioning. The mean age of diagnosis of PNETs is 29 to 38 years, and unlike MEN1, most of these lesions are solitary as opposed to multifocal.




Classification and staging


Originally referred to as islet cell tumors, PNETs were renamed as such in 2010 by the World Health Organization (WHO). Over the past 10 years, groups such as WHO, American Joint Committee on Cancer (AJCC), and the European Neuroendocrine Tumor Society (ENETS) have proposed formal staging systems for PNETs. All systems make a basic distinction between well-differentiated tumors, which tend to be more indolent, and poorly differentiated tumors, which tend to behave more aggressively.


The 2010 WHO classification of PNETs is based on the proliferative index of the tumor as measured by the Ki-67 index and mitotic count ( Table 1 ). This classification separates well-differentiated lesions into low-grade (G1) and intermediate-grade (G2) NETs, whereas all poorly differentiated lesions (G3) are labeled as neuroendocrine carcinomas.



Table 1

2010 WHO grading system for PNETs



























Differentiation Grade Ki-67 Index Mitotic Count Nomenclature
Well differentiated Low (G1) <3% <2 per 10 HPF Neuroendocrine tumor, grade 1
Intermediate (G2) 3%–20% 2–20 per 10 HPF Neuroendocrine tumor, grade 2
Poorly differentiated High (G3) >20% >20 per 10 HPF Neuroendocrine carcinoma

Abbreviations: HPF, high-power field; WHO, World Health Organization.

Data from Bosnan FT, Carneiro F, Hruban RH, et al. WHO classification of tumours of the digestive system. Lyon (France): International Agency for Research on Cancer; 2010; and Klimstra DS, Modlin IR, Coppola D, et al. The pathologic classification of neuroendocrine tumors: a review of nomenclature, grading, and staging systems. Pancreas 2010;39(6):707–12.


With regard to staging, WHO endorses staging NETs based on the TNM classification; however, the AJCC manual and ENETS differ in their use of the TNM criteria. Whereas the AJCC manual stages PNETS using the same system developed for pancreatic adenocarcinoma, the ENETS proposal is modified specifically for PNETs and incorporates a different T-stage definition. The AJCC and ENETS staging systems are summarized in Tables 2 and 3 .



Table 2

American Joint Committee on Cancer staging system for pancreatic cancer (used for both endocrine and exocrine tumors)














































Stage Tumor Node Metastasis Description
1A T1 N0 M0 Tumor limited to pancreas and size <2 cm
1B T2 N0 M0 Tumor limited to pancreas and size >2 cm
IIA T3 N0 M0 Tumor extends beyond pancreas but without involvement of the celiac axis of the superior mesenteric artery
IIB T1–T3 N1 M0 Any tumor without artery involvement that has regional lymph node metastases
III T4 Any N M0 Tumor involves the celiac axis of the superior mesenteric artery
IV Any T Any N M1 Any tumor with distant metastases

Data from Exocrine and endocrine pancreas. AJCC cancer staging manual. New York: Springer; 2010. p. 241–9.


Table 3

European Neuroendocrine Tumor Society staging system for pancreatic neuroendocrine tumors














































Stage Tumor Node Metastasis Description
I T1 N0 M0 Tumor limited to the pancreas and size <2 cm
IIA T2 N0 M0 Tumor limited to the pancreas and size 2–4 cm
IIB T3 N0 M0 Tumor limited to the pancreas and size >4 cm or invading duodenum or bile duct
IIIA T4 N0 M0 Tumor invading adjacent organs
IIIB Any T N1 M0 Any tumor that has regional lymph node metastases
IV Any T Any N M1 Any tumor with distant metastases

Data from Rindi G, Klöppel G, Alhman H, et al. TNM staging of foregut (neuro)endocrine tumors: a consensus proposal including a grading system. Virchows Arch Int J Pathol 2006;449(4):395–401.


Although both the AJCC and the ENETS staging systems appear to be independent predictors of survival among patients with PNETs, one large European cohort study found the ENETS system to provide more equally populated risk groups and more accurate predictive ability.




Pathophysiologic mechanisms


Although functioning PNETs may secrete hormones that are associated with the endocrine pancreas (islets of Langerhans), recent investigation has shown that PNETs actually arise from pluripotent stem cells of the exocrine (acinar/ductal) pancreas. These pluripotent cells then develop into atypical accumulations with both an exocrine and endocrine phenotype. The most common driver mutations implicated in PNET tumorigenesis are MEN1 , DAXX/ATRX , and mammalian target of Rapamycin ( mTOR ), found in 44%, 43%, and 14% of PNETs, respectively.


MEN1 is a tumor suppressor gene that codes for the protein Menin. Loss of function of the MEN1 gene may be inherited in an autosomal dominant fashion ( MEN1 syndrome) or be acquired as a sporadic mutation. Regardless, mutation of MEN1 is thought to be a proximal event in PNET tumorigenesis. DAXX and ATRX function together as a dimer and are involved in chromatin remodeling and stabilization. Specifically, these proteins are associated with abnormal telomeres through a telomerase-independent telomere maintenance mechanism named ALT (alternative lengthening of telomeres). Loss of DAXX or ATRX appears to result in reduced time of relapse-free and tumor-associated survival, and may signify a more aggressive subtype of PNET. The mTOR pathway is involved in cell survival and proliferation. Abnormal activation of mTOR is common in patients with PNETs, and mTOR inhibitors are a current focus of targeted molecular therapy.




Clinical presentation


The clinical presentation of PNETs is frequently divided into 2 broad categories: functioning and nonfunctioning tumors. Functional tumors present with a defined clinical syndrome secondary to hormone hypersecretion. Functional PNETs included insulinomas, gastrinomas, glucagonomas, VIPomas, and somatostatinomas. Nonfunctional tumors, which compromise 60% to 90% of all PNETs, do not present with symptoms related to hormone hypersecretion, although in some cases circulating hormone levels may be elevated nonetheless.


Functional Tumors


Functional tumors represent the minority of PNETs, and the presenting symptoms depend on the hormone that is overproduced. Insulinomas are the most common symptomatic PNET. The Whipple triad is the diagnostic hallmark of insulinomas, which consists of documented hypoglycemia (plasma glucose <50 mg/dL), neuroglycopenic symptoms (confusion, visual change, diaphoresis, tremulousness), and rapid reversal of symptoms with administration of glucose. Approximately 90% of patients exhibit the Whipple triad at the time of diagnosis. Patients generally present in the fifth decade, with a higher incidence in men. Gastrinomas are the second most common functional PNET. Ectopic pancreatic or duodenal gastrin hypersecretion leads to symptoms of ZE syndrome, which are reflux, refractory peptic ulcer disease, and secretory diarrhea. Glucagonomas, somatostatinomas, and VIPomas are rare and represent fewer than 10% of all functional PNETs. The most common presenting symptom of glucagonoma is a classic rash, migratory necrolytic erythema, which is present in 52% of patients. Surprisingly, diabetes develops in only 20% of patients before glucagonoma diagnosis. Symptoms of glucagonoma are often summarized as the “4Ds”: dermatitis, diabetes, depression, and deep vein thrombosis. VIPomas often present with large-volume, watery diarrhea that is secretory in nature and may lead to dehydration and hypokalemia. Excess VIP secretion also may lead to flushing and electrolyte abnormalities, such as hyperglycemia, hypercalcemia, and hypochlorydria. Somatostatinomas comprise fewer than 5% of PNETs, and are characterized by the clinical syndrome of diabetes mellitus, gallbladder disease, diarrhea, weight loss, and steatorrhea. Nevertheless, it is exceedingly rare for patients to present with all of these features. In most cases, somatostatinomas are diagnosed by chance after a gastrointestinal NET demonstrates immunohistochemical staining for somastostatinlike immunoreactivity.


Nonfunctional Tumors


Nonfunctional tumors represent 60% to 90% of all PNETs. As they do not present clinically with a hormonal syndrome, nonfunctional PNETs are often diagnosed later in the course of disease when they present with symptoms of compression or metastatic disease. The most common presenting symptoms of nonfunctional PNETs are abdominal pain, weight loss, anorexia, and nausea. Patients may also infrequently present with obstructive jaundice, mimicking the presentation of pancreatic adenocarcinoma.




Diagnostic evaluation


Laboratory Evaluation


When patients present with clinical syndromes suspicious for PNETs, elevated levels of the hormone in question generally confirm the diagnosis. Glucagonoma, VIPoma, and somatostatinoma are associated with elevated serum levels of glucagon, VIP, and somatostatin, respectively. Insulinoma is diagnosed by measuring elevated serum insulin levels in the setting of hypoglycemia, which is often done after a prolonged, 48-hour to 72-hour fast. Plasma glucose, C-peptide, and proinsulin also should be measured to rule out surreptitious insulin use. Gastrinoma is associated with an elevated gastrin level in 99% of cases; however, this finding may be nonspecific. Patients with hypochlorhydria from atrophic gastritis/pernicious anemia, chronic Helicobacter pylori infection, or those on proton-pump inhibitors (PPIs) may also have physiologically elevated levels of gastrin. As such, inappropriately elevated gastrin levels such as those associated with gastrinoma/ZE syndrome often require confirmation with a secretin stimulation test, which has a sensitivity of 94% and a specificity of 100%. PPIs should be stopped for a week before these tests.


Several tumor markers have been studied with regard to PNETs. The most frequently used is chromogranin A (CgA), which is an acidic glycoprotein that is produced and excreted by granules within NETs. CgA at a cutoff of 60.7 ng/mL has a sensitivity of 77% and a specificity of 56% for the diagnosis of PNETs. However, more recent analysis suggests that when insulinomas (which do not produce as much CgA) are excluded, a cutoff of 74 ng/mL has a specificity of as high as 91%. A nomogram has recently been proposed to predict tumor burden, evaluate response, and predict survival for patients with well-differentiated PNETs and liver metastases based on CgA levels. Other tumor markers proposed include pancreatic polypeptide and neuron specific enolase; however, their use in routine clinical practice remains less well defined. Like CgA, all currently available biomarkers suffer from significant false-positive results.


Imaging


Several imaging modalities exist to aid the clinician in the localization and staging of PNETs. Computed tomography (CT), MRI, endoscopic ultrasound (EUS), somatostatin-receptor scintigraphy (SRS), and functional PET (fPET) are all commonly used in clinical practice. Arterial stimulation with venous sampling is a technique reserved for functional tumors that cannot be located with conventional techniques.


Computed tomography


CT scans are readily available, noninvasive tests often used in the diagnosis and staging of PNETs. PNETs are highly vascular lesions best appreciated with triple phase-contrast studies, as they enhance during the arterial phase with washout during the portal venous phase. The sensitivity of CT scan for detecting PNETs ranges from 63% to 82% and the specificity ranges from 85% to 100%. Two particular limitations of CT are the detection of lesions smaller than 2 cm and the detection of insulinomas. For these lesions, EUS is the imaging modality of choice. The enhancement characteristics of a PNET on CT may also have prognostic value, as lesions that hypoenhance on CT scan tend to have worse overall survival than lesions that are isoenhancing or hyperenhancing (5-year survival 54% vs 89% vs 93%).


MRI


MRI is frequently a modality of choice to evaluate pancreatic disease. The sensitivity of MRI for the detection of PNETs ranges from 85% to 100%, with a specificity from 75% to 100%. Similar to CT, MRIs are best appreciated in the late arterial enhancement phase, and the signal is low in T1-weighted images and high in T2-weighted images. Specific MRI characteristics also may have prognostic value, as a study of 18 patients demonstrated that the apparent diffusion coefficient in diffusion-weighted MRI correlated with the Ki-67 labeling index.


Endoscopic ultrasound


EUS is increasingly used in the evaluation of PNETs ( Fig. 1 ). By combining endoscopy and ultrasound, EUS is not only able to localize and stage PNETs, but also to confirm and, in rare cases, treat disease. EUS-guided fine-needle aspiration (EUS-FNA) is a reliable means of obtaining tissue for diagnosis, and EUS-guided injection of ablative agents, such as ethanol, has proven successful in the treatment of insulinomas in case reports. EUS is superior to CT in the detection of smaller lesions, and can visualize PNETs as small as 2 to 5 mm. Furthermore, EUS is particularly useful in the detection of insulinomas that are often multifocal, and may lack somatostatin receptors and are consequently not well visualized on SRS. Limitations of EUS for the detection of PNETs include lesions in the pancreatic tail; this technique is also highly operator dependent. Overall, the sensitivity and specificity for EUS in the detection of PNETs ranges from 82% to 93% and 92% to 95%.




Fig. 1


EUS image of well-circumscribed 10.1 × 8.0-mm hypoechoic pancreatic head mass with peripheral enhancement consistent with NET. Pathology revealed well-differentiated, low-grade (G1) PNET.

( Courtesy of Michelle Kang Kim, MD, PhD, New York, NY.)


Somatostatin-receptor scintigraphy


Similar to MRI, SRS is often used to localize PNETs when traditional cross-sectional imaging fails. SRS uses radiolabeled somatostatin analogues that bind to receptors expressed by the PNET. Hence, lesions with few somatostatin receptors, such as insulinomas, are often missed with this technique. The standard method of performing SRS is with injection of 111In-DTPA-octreotide (OctreoScan). SRS is particularly useful in assessing the burden of extrapancreatic metastatic disease, and also should be routinely performed to assess the somatostatin receptor status before treatment with somatostatin analogues. Overall, the sensitivity of SRS to detect gastrinomas, VIPomas, glucagonomas, and nonfunctioning PNETs ranges from 75% to 100%; however, for insulinomas, the sensitivity falls to 50% to 60%.


PET


The role of PET scanning in the detection and staging of PNETs is evolving. Although traditional PET imaging with 18F-Fluorodeoxyglucose (FDG) is not useful for most PNETs, the emerging use of PET imaging with 68Ga-labeled somatostatin analogues and hybrid imaging with PET/CT have proven useful. In a recent study of 23 PNETs, the diagnostic sensitivity of PET/CT with the somatostatin analogue 11-C-5-HTP was 96%, compared with 46%, 77%, and 68% for SRS, SRS/CT, and CT alone, respectively. However, because of limited availability of functional tracers, these new PET techniques are more widely available in Europe, and less routinely performed in North America. As such, functional and hybrid PET is generally reserved for cases in which conventional techniques fail or show contradictory results.




Treatment


In general, surgery remains the cornerstone of treatment for PNETs and is the only curative option. Sporadic, nonmetastatic disease should be resected regardless of functional status. Depending on the location of the lesion, options include enucleation, central pancreatectomy, pancreaticoduodenectomy, and distal pancreatectomy. Metastatic disease requires a multidisciplinary approach. Isolated liver metastases are treated aggressively with either surgery or local ablative or embolic therapies. For widespread metastatic disease, surgery remains an option provided that at least 90% of the tumor burden can be resected. The approach to nonresectable disease is based on the degree of differentiation of the lesion. Well-differentiated (G1 or G2) lesions are indolent and therapy is focused on symptom control. Options include somatostatin analogues, systemic chemotherapy, targeted molecular therapy, and peptide receptor radiotherapy. Poorly differentiated (G3) lesions respond well to chemotherapy but are associated with a more rapid deterioration and worse survival. The National Comprehensive Cancer Network has recently published clinical practice guidelines on the evaluation and treatment of PNETs.


Surgery


Surgical resection remains the treatment of choice for sporadic, nonmetastatic disease; however, among patients with PNETs associated with a genetic syndrome, such as MEN1, VHL, NF1, or TSC, the role of surgery is more controversial.


Sporadic lesions


As mentioned previously, approximately 90% of PNETs are sporadic. For these lesions, surgery is the treatment of choice. Pancreaticoduodenectomy, distal pancreatectomy, and more parenchymal-preserving procedures, such as enucleation and central pancreatectomy, have all been used to treat PNETs. The choice of technique relies on several factors, including location of the lesion, malignant potential, functional status, and the presence of distant metastases. Although surgery may be performed for curative intent (as in the case of solitary, sporadic lesions), there is also a role for palliative debulking of the metastatic burden of PNETs, especially in the presence of functional tumors causing symptoms of hormonal excess.


Enucleation


Pancreatic enucleation is a parenchymal-sparing procedure that results in much lower rates of endocrine and exocrine insufficiency than traditional pancreatic resection. Previously, this procedure was reserved for small, low-grade PNETs; however, more recently enucleation is being considered as an alternative to standard resection in carefully selected patients. One major limitation of this procedure is that it does not allow for adequate lymph node sampling, which may limit the potential of complete oncological resection in certain patients. Nevertheless, comparable data are now emerging suggesting that enucleation should be the procedure of choice for small benign and premalignant lesions when technically appropriate. A recent meta-analysis of 13 studies comprising 1101 patients found that compared with standard resection for small pancreatic lesions, enucleation was a shorter procedure that resulted in less blood loss and less pancreatic insufficiency despite comparable rates of mortality and complications. The most common complication associated with this procedure, pancreatic fistula, however, was higher in the enucleation group (odds ratio 1.99, confidence interval [CI] 1.2–3.4, P <.01). Nevertheless, pancreatic fistulae that result from enucleation tend to be less severe than those following standard resection. With regard to long-term survival, 5-year and 10-year survival appears no different between enucleation and standard resection.


Central pancreatectomy


Central pancreatectomy is another parenchymal-preserving procedure that may be indicated for lesions in the pancreatic neck or body, close to the pancreatic duct. During this procedure, the pancreatic neck is divided, the tumor is excised, and pancreatico-jejunal or pancreatico-gastric anastomoses are created. Similar to enucleation, central pancreatectomies have been increasingly performed in recent years. Furthermore, compared with pancreaticoduodenectomy or distal pancreatectomy, central pancreatectomy is also associated with decreased operative time, less blood loss, and improved endocrine/exocrine function postoperatively with no difference in mortality. Central pancreatectomy is technically challenging and traditionally performed open; however, recent series suggest promising results from laparoscopic and robot-assisted approaches. Because central pancreatectomy does not allow access to adequate lymph node sampling, certain groups recommend it only for cases that would otherwise be appropriate for enucleation, but are not amenable due to location deep in the pancreatic parenchyma.


Pancreaticoduodenectomy and distal pancreatectomy


Pancreaticoduodenectomy and distal pancreatectomy are the more traditional surgical approaches to PNETs. Pancreaticoduodenectomy is indicated for lesions in the head of the pancreas, whereas distal pancreatectomy is used for lesions in the body or tail of the pancreas. Distal pancreatectomy may or may not be performed with concurrent splenectomy. Although splenic preservation is associated with decreased risk of infectious and severe complications, it is often reserved for more benign lesions like insulinomas, in which risk of malignancy is low and adequate lymph node harvest is less important.


Treatment of hepatic metastases


The surgical approach to metastatic disease divides patients into 2 categories: those with potentially resectable disease (often isolated liver metastases), and those with more extensive, unresectable disease. In general, the approach to patients with potentially resectable disease is aggressive. For isolated liver metastases, partial hepatectomy and less invasive methods, such as transarterial chemoembolization (TACE), radiofrequency ablation (RFA), yttrium-90 radioembolization (RE), cryotherapy, and microwave coagulation have all been used with varying degrees of success. An aggressive surgical approach for metastatic and locally advanced tumors is associated with increased long-term survival.


Among patients with advanced PNETs and a large metastatic burden, surgery also may be considered even if complete resection is not possible. Although recurrence is expected, debulking is associated with improved symptom control and prolonged survival. A recent retrospective study of 72 patients with nonfunctioning metastatic PNETs found a 1-year and 5-year disease-free survival rate of 58.1% and 3.5% for patients who underwent greater than 90% debulking of the tumor burden, similar to 53.7% and 10.7% for those who underwent complete resection. With regard to symptom control, a separate series of 170 patients who underwent major debulking of hepatic metastases achieved a 96% rate of partial or complete response of symptoms, although this number included both ileal carcinoids and PNETs. This improvement in symptom control may justify the high rate of recurrence of 88%, and the general consensus is that there is a role for palliative debulking as long as 90% of the disease burden can be resected.


Multiple Endocrine Neoplasia Type I–Associated Lesions


The role of surgery for PNETs associated with MEN1 is less well defined, particularly for gastrinomas. For small gastrinomas, a prospective study of 81 patients with MEN1-associated gastrinomas found that patients with tumors smaller than 2.5 cm who did not undergo surgical resection had equivalent 15-year survival to those patients with tumors larger than 2.5 cm who did have surgery (89%–100%). Hence, these small lesions can likely be observed without exposing the patient to the morbidity and mortality of pancreatic resection. For MEN1-associated gastrinomas larger than 2.5 cm, resection of solitary pancreatic lesions has been associated with improved long-term survival. However, because MEN1-associated gastrinomas are often multiple and may be located in the duodenum, complete surgical resection may necessitate a formal pancreaticoduodenectomy. In these cases, expert opinion differs regarding medical versus surgical management.


Insulinomas, glucagonomas, VIPomas, and somatostatinomas that arise in the background of MEN1 should be resected, as malignant PNETs constitute the leading cause of death among patients with MEN1. For insulinomas, those associated with MEN1 are often multiple, small, and may evade conventional localization techniques. Hence, although these lesions should be resected, no definitive consensus exists regarding the optimal technique. Given that local resection has a high chance of failure, some advocate subtotal pancreatectomy with enucleation of lesions in the head of the pancreas.


For nonfunctioning PNETs associated with MEN1, current guidelines suggest conservative management of lesions smaller than 1 cm, and resection of lesions larger than 2 cm or those that display rapid growth such as doubling in size over a 3-month to 6-month period. Lesions 1 to 2 cm in size may be managed conservatively or surgically.


Systemic Therapy


Although the only curative treatment for PNETs is surgery, several medical therapies exist for those with unresectable metastatic disease resulting in tumor bulk or hormonal excess. Well-differentiated and poorly differentiated PNETs respond differently to medical management, and therefore should be approached differently.


Well-differentiated pancreatic neuroendocrine tumors


Well-differentiated PNETs are either low-grade (G1) or intermediate-grade (G2) NETs. These tumors are more indolent, and medical management is generally focused on symptom control. For functioning tumors, initial therapy includes treating the manifestations of hormonal excess by using somatastatin analogues (SSAs) when appropriate. Systemic chemotherapy, targeted molecular therapy, and peptide receptor radiotherapy may be used for refractory disease and nonfunctioning tumors.


Hormonal excess and somatostatin analogues


Conventional treatment of symptoms should be used whenever appropriate. For example, gastrinomas should be treated with high-dose PPI therapy, and insulinomas should be treated with carbohydrates and diazoxide. For refractory cases and other PNETs like VIPomas, somatostatinomas, and glucagonomas, SSAs, such as octreotide or lanreotide, may be used. SSAs bind to somatostatin receptors expressed by most NETs and subsequently reduce the secretion of hormonally active peptides. SSAs are generally less useful in insulinomas, as far fewer of these tumors express the somatostatin receptor. Furthermore, symptoms of hypoglycemia may transiently worsen with concurrent suppression of glucagon secretion.


In addition to controlling the symptoms of hormonal excess, SSAs may also limit tumor growth. Most recently, a multinational study of well-differentiated and moderately differentiated metastatic enteropancreatic NETs comparing lanreotide with placebo found significantly improved progression-free survival for patients taking lanreotide (65.1% vs 33.0% at 24 months, hazard ratio (HR) 0.47, 95% CI 0.30–0.73). This result is consistent with previous studies examining tumor regression with SSAs in gastropancreatic NETs. SSAs are generally well-tolerated, with the most common side effects being diarrhea/loose stools from pancreatic insufficiency, abdominal discomfort, nausea, and the development of gallstones.


Chemotherapy


Several chemotherapeutic regimens have been used to control the metastatic burden of PNETs. The initial standard therapy was with streptozocin-based regimens, but side effects, such as kidney injury, hair loss, and nausea, have limited the widespread adoption of such treatments. Instead, temozolomide and oxaliplatin-containing regimens are more often used. Temozolomide is an oral alkylating agent whose efficacy appears to be dependent on expression of the DNA repair enzyme O(6)-methylguanine-DNA methyltransferase (MGMT). PNETs deficient in MGMT respond better to temozolomide-based therapy than those with intact expression. Oxaliplatin is a platinum-based agent that when combined with capecitabine has shown efficacy in the treatment of well-differentiated NETs after progression with somatostatin analogues.


Targeted molecular therapies


Targeted molecular therapies for PNETs include those that target the vascular endothelial growth factor (VEGF) and mTOR pathways. Tyrosine kinase inhibitors, such as sunitinib, sorafenib, and pazopanib, as well as the direct VEGF blocker bevacizumab work via the VEGF pathway. Everolimus and temsirolimus inhibit the mTOR pathway.


Tyrosine kinase inhibitors


Sunitinib is currently the only tyrosine kinase inhibitor approved by the Food and Drug Administration (FDA) for the treatment of unresectable, well-differentiated PNETs. Sunitinib is an orally available medication that targets the VEGF receptor pathway. A large randomized, phase III trial reported a median progression-free survival of 11.4 months for patients treated with sunitinib versus 5.4 months for placebo (HR 0.42, 95% CI 0.25–0.66, P <.001). Sorafenib and pazopanib, although not FDA approved, have also shown promise in the treatment of PNETs. In a phase II trial of unresectable gastropancreatic NETs, which included 43 PNETs, progression-free survival was reported in 60.8% of evaluable patients with PNET. However, 43% of patients experienced significant grade 3 or 4 toxicity.


Mammalian target of Rapamycin inhibitors


Everolimus was FDA approved in 2011 for the treatment of unresectable, locally advanced, or metastatic PNETs based primarily on data from 2 large, phase III trials. RADIANT-2 compared everolimus plus octreotide with placebo plus octreotide in 429 patients with advanced, functionally active NETs. Patients treated with everolimus plus octreotide had improved median progression-free survival (PFS) (16.4 vs 11.3 months, P = .026). RADIANT-3 compared daily everolimus 10 mg with placebo in 410 patients specifically with PNETs. Similarly, the everolimus-treated patients had an increased median PFS compared with placebo (11.0 vs 4.6 months, P <.001). Temsirolimus was evaluated in a phase II trial of 37 patients with advanced, progressive, neuroendocrine carcinoma ; however, only a modest intention-to-treat response rate of 5.6% was noted. Hence, further studies investigating the use of temsirolimus as monotherapy for advanced NETs were deferred.


Bevacizumab


Bevacizumab is a monoclonal antibody to circulating VEGF-A that inhibits angiogenesis and is currently used to treat a number of different malignancies. With regard to NET, bevacizumab has been studied in conjunction with several other therapies, including everolimus, temsirolimus, and FOLFOX, with promising results. Nevertheless, it is not FDA approved for the treatment of PNET and its use remains investigational. An open-label, phase II study also combined bevacizumab with sorafenib; however, despite a median PFS of 12.4 months, 63% of patients experienced a grade 3 or 4 adverse reaction.


Peptide receptor radiotherapy


Peptide receptor radiotherapy (PRRT) remains an investigational procedure in the United States, despite extensive experience in Europe. PRRT couples radioactive isotopes, such as yttrium ( 90 Y) and lutetium ( 177 Lu), to SSAs to deliver targeted radiotherapy to NET cells. Early results are promising. The largest series of yttrium-coupled octreotide (90Y-DOTA-TOC) comprising 1109 patients with metastatic gastropancreatic NETs and disease progression treated with a median of 2 cycles of therapy found 34% of patients achieved radiographic response, 15% had biochemical response, and 29.7% had symptomatic improvement. The median survival from diagnosis was 94.6 months and the rates of transient grade 3 or 4 toxicity or permanent grade 4 or 5 toxicity were 12% and 9%, respectively. A second study of 504 patients treated with Lu-177 reported similar efficacy but a much lower toxicity rate of 3.6%.


Poorly differentiated pancreatic neuroendocrine tumors


Poorly differentiated PNETs are high-grade (G3) tumors that are referred to as neuroendocrine carcinomas. In general, poorly differentiated NETs respond better to chemotherapy, but are associated with a more rapid deterioration and worse survival. Platinum-based chemotherapeutic regimens such as cisplatin and etoposide have shown promise in these lesions; however, prognosis remains poor. A study of 36 patients with either poorly differentiated NETs or a rapidly progressing clinical course treated with cisplatin/etoposide showed a radiographic response rate of 36%, but a median survival time of only 19 months. Few data exist for second-line therapy, and no studies have looked at PNETs specifically. Furthermore, little is known about the potential role of targeted molecular therapies or peptide receptor radiotherapy among patients with poorly differentiated PNETs.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Sep 6, 2017 | Posted by in GASTROENTEROLOGY | Comments Off on Islet Cell Tumors of the Pancreas

Full access? Get Clinical Tree

Get Clinical Tree app for offline access