Review. Pancreatic endocrine tumors: Recent advances

Annals of Oncology 10 Suppl. 4: S170-S176, 1999. 1999 Kluwer Academic Publishers. Printed in the Netherlands. Review Pancreatic endocrine tumors: Recent advances R. T. Jensen Digestive Diseases Branch, National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health, Bethesda, MD USA Summary Pancreatic endocrine tumors (PETs) can be divided on a clinical and pathologic basis into ten classes [insulinomas, gastrinomas (Zollinger-Ellison syndrome), VIPomas (Verner-Morrison syndrome, WDHA, pancreatic cholera), glucagonomas, somatostatinomas, ACTH-releasing tumors (ACTHomas), growth hormone-releasing factor secreting tumors (GRFomas), nonfunctioning or pancreatic polypeptide secreting tumors (nonfunctioning PET), PETs causing carcinoid syndrome and PETs causing hypercalcemia)]. Recent reports suggest calcitoninsecreting PETs also rarely occur but whether they cause a distinct clinical syndrome is unclear. PETs resemble carcinoid tumors histologically; in their ability to synthesize and frequently secrete multiple peptides such as neuroendocrine cell markers (chromogranins); their biologic behavior and their tumor growth patterns. Both groups of tumors are highly vascular, have high densities of somatostatin receptors and similar tumor localization studies including somatostatin receptor scintigraphy are used for both. PETs, similar to carcinoids causing the carcinoid syndrome, require two separate treatment options be considered: treatment Introduction Pancreatic endocrine tumors (PET's) share cytochemical properties with carcinoid tumors, medullary thyroid cancer, melanomas and pheochromocytomas and have been collectively called APUDomas (an acronym for amine precursor uptake and decarboxylation). PET's, similar to malignant carcinoid tumors, frequently synthesize and secrete multiple peptides, some of which are biologically active causing distinct clinical syndromes (Table 1). PET's resemble carcinoids and other ADUDomas in producing products characteristic of neuroendocrine differentiation such as neuron specific enolase, chromogranins (A,B,C) and synaptophysin. The clinical management of these tumors presents a number of unique challenges because all PETs except insulinomas are frequently malignant (Table 1). Therefore, treatment options must be directed both at the tumoral process ger se as well as the clinical syndrome caused by the ectopically produced hormone (Table 1). Recently there have been a number of insights into the natural history of some PETs which allowed identification of prognostic factors and which will allow treatment options to be more carefully planned, as well as their genetic basis, newer methods to diagnose PETs, and improved methods to localize PETs which allow better staging. Furthermore, important insights into the molecular basis for PETs and carcinoids has been obtained. In this short review a number of these advances will be briefly summarized. Because a directed against the hormone-excess state and treatment directed against the tumor per se because of their malignant nature. In the last few years there have been advances in tumor diagnosis, localization methods, treatment approaches particularly related to the use of synthetic somatostatin analogues, and the definition of the role of surgical procedures in these diseases. Important other advances include insights into the long-term natural history of PETs particularly from studies of gastrinomas, which allow prognostic factors to be identified and the timing of treatment options to better planned, as well as insights into the molecular basis of these disorders. The latter includes both a description of the molecular basis of the genetic inherited syndromes associated with PETs or carcinoid tumors, as well as an increased understanding of the molecular basis for sporadic PETs or carcinoid tumors. Each of these areas will be briefly highlighted in this presentation. Key words: apudoma, carcinoid, gastrinoma, glucagonoma, GRFoma, insulinoma, MEN1, neurofibromatosis, nonfunctional pancreatic endocrine tumor, somatostatinoma, VIPoma, von Hippel Lindau disease. number of other presentations in this mini-symposium deal with advances in specific areas of treatment of these tumors including: the use of somatostatin receptor scintigraphy and somatostatin analogues for treatment; intraoperative staging and the role of surgery in their treatment; the treatment of advanced disease by chemotherapy and interferon; and immunotherapy and genetic therapy in their treatment, these areas will be dealt with in this review only in the context of the other areas discussed. Current classification of PET's There continues to be no absolute agreement on the best name for this group of diseases or on their classification. These tumors are frequently called islet cell tumors, however, it is unproven they originate from pancreatic islets (Table 1). The term pancreatic endocrine tumor (PET) is frequently used, however it is also a misnomer, because many of these tumors can occur outside the pancreas [gastrinomas, VIPomas, GRFomas, ACTHomas, somatostatinomas]. However, because this term is widely used it will be retained in this chapter. PETs are characteristically classified by the type of clinical syndrome they cause and therefore can be divided into functional or nonfunctional. Many nonfunctional PETs by this classification are not truly nonfunctional because they frequently secrete hormones and peptides [neurotensin,

171 Table I. Pancreatic endocrine tumor (PET) syndromes Name Biologically active peptide secreted Incidince (new cases/10 6 population/year) Tumor location Malignant Associated (%) with MEN-I Patients with MEN-I Develop I Established specific functional syndrome Zollinger-Ellison syndrome Gastrin 0.5-1.5 Duodemim-70% Pancreas-25% Other sites-5% 60-90 20-25 54 Insulinoma Insulin 1-2 Pancreas (>99%) <10 4-5 21 VIPoma [Verner-Morrison syndrome, pancreatic cholera, WDHA] Vasoactive intestinal peptide 0.05-0.2 Pancreas-90%, adult Other-10%, neural, adrenal, periganglionic 40-70 6 17 Glucagonoma Glucagon 0.01-0.1 Pancreas (100%) 50-80 1-20 3 Somatostatinoma GRFoma ACTHoma PET causing carcinoid syndrome PET causing hypercalcemia Somatostatin Growth-hormone releasing hormone ACTH Serotonin? tachykinins II Possible specific functional syndrome PET secreting calcitonin III No functional syndrome Ppoma/ Nonfunctional Calcitonin None Unkown (43 cases) Pancreas-55% Duod/Jej-44% Pancreas-30% Lung-54% Jejunum-7%, Other 13% Pancreas pancreatic polypeptide, chromogranins, a-subunit of human chorionic gonadotropin, neuron specific enolase], however, these cause no distinct clinical syndrome. For example, recent studies show that similar to carcinoid tumors, PETs not associated with a distinct clinical syndrome, (i.e. nonfunctional PETs), secrete chromogranin A in 60%-90% of patients [ 1-4]. The term nonfunctional PET is widely used it will be retained in this review to mean any PET either secreting no products or secreting products not causing any specific clinical syndrome. The functional PETs can be classified into nine distinct syndromes (Table 1). In addition to the 6 syndromes generally included (gastrinomas, insulinomas, VIPomas, glucagonomas, GRFomas, somatostatinomas), there are sufficient numbers of wellcharacterized cases which have been recently reviewed to include in this classification pancreatic ACTHomas [5], PETs causing the carcinoid syndrome [6,7], and PETs causing hypercalcemia [5,8]. A recent review of 6 cases proposes including PETs secreting calcitonin [9] because one-half of the patients had diarrhea, which disappeared with the treatment of the tumor in two of the patients. The proposal that the diarrhea is due to the calcitonin secreted by 1-2 Pancreas (<1% all carcinoids) Pancreas (rare cause hypercalcemia) Pancreas (rare cause hypercalcitonema) Pancreas (100%) >70 >60 >95 60-88 84 >80 >60 45 16 16 18-44 80-100 the PET is supported by the finding that 25%-42% of patients with medullary thyroid cancer with hypercalcitonemia developed diarrhea, likely secondary to a motility disorder [10]. However, this PET is classified in Table 1 as a possible specific syndrome because only a few cases have been well studied. The tenth well-established PET syndrome is that due to nonfunctional PETs where all of the symptoms are due to the tumor rjer se (pain, jaundice, etc.) and not to the secreted product(s) (Table 1). Advances in diagnosis of PET's Because the PET's often have an indolent course and can masquerade as more common other conditions, the delay in their diagnosis continues to be 4-6 years. For example, a recent study reports even in patients with MEN1 where gastrinomas are known to be the most common functional PET, the delay in diagnosis remains 4.1 years [11] Only by having a high index of suspicion and by demonstrating an elevated plasma level of the appropriate hormone with a simultaneous altered clinical state (hyperacidity,

172 hypoglycemia, increased stool output, acromegaly, etc.) can the proper diagnosis be established. Each of syndromes listed in Table 1 except nonfunctional PET's is a clinical syndrome and requires clinical evidence for a hormone excess state. Strictly speaking immunocytochemistry alone is not sufficient to make these diagnoses. Many PETs synthesize multiple peptides which can be demonstrated immunocytochemically but are not increased in the plasma and are not causing the appropriate clinical syndrome. The functional PETs continue to be frequently misdiagnosed because the diagnosis is based only on immunocytochemical results. Somatostatinomas are especially often misdiagnosed with case reports based only on immunocytochemical results [12,13]. Two recent developments may have an important effect on the diagnosis of nonfunctional PETs, which are almost always diagnosed late in their disease course [14]. The first is the establishment in numerous studies that plasma chromogranin A is elevated in 60%-100% of both functional and nonfunctional PETs as well as carcinoid tumors [1-4]. Furthermore, recent studies demonstrate serum chromogranin levels are useful in an individual patient with a carcinoid tumor or PET in assessment tumor progression, relapse and tumor burden [2-4]. In the case of nonfunctional PETs and carcinoid tumors the use of serum chromogranin levels may prove to be an important, non-imaging method that is cost effective to diagnose these tumors as well as follow tumoral changes post treatment. A second important advance in the diagnosis of PET's is the proposed use of somatostatin receptor scintigraphy (SRS) for this purpose. Recent studies demonstrate PET's as well as carcinoids and other APUDomas possess high densities of somatostatin receptors [15,16]. Using [ lu In-DTPA-DPhe']octreotidefor scintigraphy, SRS has proven to be the most sensitive modality for imaging PETs and carcinoids [15-17]. A recent study [18] demonstrates SRS may also be useful for differentiating PET's from pancreatic adenocarcinomas and therefore useful in diagnosis of the PET. In this study [18] 65% of all patients with suspected PETs demonstrated positive SRS localization and 0 of 26 patients (0%) with pancreatic adenocarcinomas. Furthermore, 5 of 12 patients (42%) with long-term survival (>3 years) with a diagnosis of pancreatic adenocarcinoma, had positive SRS localization demonstrating this proportion of patients actually had nonfunctional PETs misdiagnosed as adenocarcinoma [18]. If SRS is used for diagnosis of a pancreatic lesion as a possible PET, it will be important that the result be interpreted carefully within the clinical context of the patient, because a recent prospective study [ 19] demonstrates that 12% of SRS localizations can be a false positive localization for a PET. However, if the SRS result is interpreted within the clinical context, in only 3% of cases did the SRS false positive result cause a management change [19]. Clinical aspects-recent insights Recent studies in patients with Zollinger-Ellison syndrome [ZES] have provided a number of important advances. Currently in almost every case gastric acid hypersecretion can be controlled by H + -K + ATPase inhibitors (omeprazole, pantoprazole, lansoprazole) short- and long-term. In one recent large study of 212 patients with ZES, no patients died from acid-related causes [20]. In another study [21] involving 116 patients treated for up to 10 years, acid hypersecretion was controlled medically in all cases. Chronic treatment with H + -K + ATPase inhibitors has proven safe. Whereas the possible effects of long-term H + -K + ATPase inhibitor treatment on the gastric mucosa in patients with ZES has been well-studied with no evidence of an increased risk of gastric carcinoids for up to 5 years of continuous treatment, the possible long-term nutritional effects on iron and vitamin B 12 absorption of such treatment are only recently investigated. Numerous clinical and laboratory studies demonstrate chronic hypochlorhydria/ achlorhydria can decrease iron [22] and vitamin B 12 absorption [23]. Recent studies demonstrate treatment with H + -K* ATPase inhibitors [22] in patients with ZES for up to 12.5 years had no effect on body iron stores or hematological parameters, even through 45% of the patients met at least one criteria for drug-induced acid hyposecretion. In contrast, in 131 patients with ZES with a mean duration of 4.5 years of treatment with H + -K + ATPase inhibitors and a maximum of up to 12 years, serum vitamin B, 2, but not serum folate levels were affected. Serum vitamin B, 2 levels were lower in patients treated long-term with omeprazole (P=0.03), especially those with sustained acid hyposecretion (P=0.0001) or complete achlorhydria (P <0.00001). In 68 patients evaluated 5 years apart, vitamin B, 2 levels decreased by 30% in those rendered achloryhydria by H + -K + ATPase inhibitors [23]. Eight patients (6%) developed subnormal vitamin B, 2 levels. This study [23] has important implications for chronic treatment with H + -K + ATPase inhibitors in ZES and raises the possibility that other non-zes patients treated chronically with these drugs may also be at risk for developing vitamin B 12 deficiency. Recent studies demonstrate that the most common location for gastrinoma is not the pancreas, but instead duodenal tumors are at least 3 to 4 times more common [24,25]. This difference from older studies is likely explained by the fact that duodenal gastrinomas were not routinely carefully sought for in older studies and only if careful inspection of the duodenum is done with performance of a duodenotomy will these small gastrinomas be found [26]. The most important clinical advances in insulinomas, VIPomas, glucagonomas, and GRFomas are related to the ability to now image these tumors with SRS as well as to treat the hormone-excess state with synthetic somatostatin analogues. Particularly important in the latter area is the development of longacting (every 10-30 days administration) somatostatin preparations (lanreotide SAR, octreotide, LAR). Because these areas are specifically covered in other presentations they will not be dealt with further here. Pancreatic ACTHomas are included in Table 1 because they account for 4%-l 6% of all ectopic Cushing's syndrome (Table 1). An important clinical observation is that 96% of the patients with this PET syndrome have a malignant tumor and therefore if careful imaging of the abdomen does not detect a pancreatic mass or liver metastases in a patient with the ectopic ACTH syndrome, it is very unlikely a pancreatic ACTHoma is present [27]. PETs causing the carcinoid syndrome are usually large and 68%-88% are malignant [6,7]. These account for <1 % of all carcinoids [6]. The carcinoid syndrome is present in 34%- 65% of these patients [6,7]. Even though foregut carcinoids

such as pancreatic carcinoids may lack dopa decarboxylase and thus the tumor cannot convert 5-hydroxytryptophan to 5 hydroxy-tryptamine (serotonin), 84%-85% of patients with pancreatic carcinoids with the carcinoid syndrome have increased urinary 5-HIAA levels which can be used for their detection. PETs causing hypercalcemia are almost all malignant (17 of 18=94%) [8] with 84% having liver metastases at diagnosis [8] and the primary is large (>5 cm in all reported cases) [8]. The serum PTH level was not elevated in any patient [8]. PTHrP was detected in the tumor in 7 of the 8 tumors (88%) examined and serum PTHrP was elevated in the 4 cases examined. In remains unclear whether PTHrP is responsible for hypercalcemia in all of these cases. Recent insights from studies of PET's natural history and identification of prognostic factors Because patients with PET's frequently died form the hormone-excess state, PET's are uncommon tumors and patients with nonfunctional PETs presented late in their course of disease (>60% have metastatic disease to the liver at diagnosis), until recently the natural history of PET's had not been systematically studied. Recent studies on patients with gastrinomas have provided a number of important insights into this area that are likely applicable to the other less common PETs. This has occurred for a number of reasons. ZES can be controlled medically in almost every case so long-term study of only the natural history of the gastrinoma is possible. Gastrinomas are malignant in 60%- 90% of older studies so they resemble all other PETs except insulinomas in their malignant potential [12,28]. Furthermore, a proportion of patients with ZES are diagnosed early before metastatic spread has occurred so patients with all stages of tumor extent are available for study. Lastly, ZES is sufficiently common that enough? 100 No liver metastases iimi.ni inn p = 0.02 -ex Single liver lobe metastases Diffuse liver metastases 5 10 15 20 YEARS SINCE DIAGNOSIS Figure 1. Effect of the presence and extent of liver metastases on survival in patients with gastrinomas. Survival rates are calculated using death due to gastrinoma-related causes as the outcome. There were 158 patients with no liver metastases of whom 6 died [solid circles], 14 patients with metastases in a single lobe of the liver of whom three died (open circles) and 27 patients with diffuse liver metastases of whom 21 died (squares). Figure modified from Yu et al [20]. 173 patients are available for systematic study. The results detailing the natural history and prognostic factors of the gastrinoma from studies of ZES are likely applicable to other less common PETs, because gastrinomas resemble these other PETs in metastatic pattern, growth patterns, histologically and in response to treatment [12]. Two recent studies from the MH [20,29] involving 185 patients and 221 patients with ZES analyzed determinants of metastatic spread, survival, disease course in patients with longstanding disease and prognostic factors. No patient in either study [20,29] died of an acid-related cause, so that the hormoneexcess state was not a factor in determining the natural history. Similar to early studies, the presence and extent of liver metastases was the most important determinant of survival (Figure 1) [20,29]. These studies [20,29] revealed that the gastrinoma in 75% of patients pursues a nonaggressive course, whereas in 25% the gastrinoma pursues an aggressive course. The 10-year survival in patients with the nonaggressive tumor course was 96% and only 30% in patients with the aggressive tumor course [29]. The factors associated with a poor prognosis are listed in Table 2 and include the presence of liver metastases (either initially or their development with time), the extent of liver metastases (diffusosingle or limited metastases in both lobes (<10) >no metastases (Figure 1); the development of bone metastases or of ectopic Cushing's syndrome; a large primary tumor (>3 cm); female gender; MEN1 absent; a short clinical course prior to diagnosis (<3 years from onset to diagnosis); a markedly increased serum gastrin level (mean in aggressive -5157 pg/ml vs 1700 pg/ml in nonaggressive); a primary tumor that was pancreatic rather than duodenal in location and the presence of lymph node metastases (Table 2). An additional study using flow cytometry to analyze changes in gastrinomas from 59 patients [30] demonstrated a high S phase, low percent nontetraploid aneuploid and high percent multiple stem line aneuploid were frequently present in aggressive gastrinomas. In the long-term study [20] with a mean follow-up of 13.8 + 0.6 yrs (range 0.4-41 yrs), 32% of the patients with ZES died, one-half from a ZES-related cause. The ZES-related causes of death were all related to the gastrinoma with the causes being: liver metastases with progressive inanition (63% of total ZES-related deaths); development of a secondary PET syndrome (30%), of which 21% were due to ectopic Cushing's syndrome; liver metastases causing hepatic failure (12%) or with secondary infection (6%) and other tumor-related causes (12%). The development of bone metastases or ectopic Cushing's syndrome were particularly predictive of a poor prognosis with patients only surviving 1.9 + 0.4 and 1.7 + 0.4 years after their diagnosis, respectively [20]. A recent study [31] demonstrated bone metastases only occur in patients with advanced ZES; were present in 31 % of patients with liver metastases and 0% with only lymph node metastases; were best detected by SRS with bone scan having lower sensitivity and specificity; and the frequency of areas of their occurrence was pelvis/sacroiliac joints>scapula> ribs. In all cases detection of bone metastases altered management [31]. The aggressive and nonaggressive pattern of growth behavior of gastrinomas also appears to apply to patients with metastatic tumors in the liver. In a recent study [32], in 19 such patients followed over time, 26% of patients demonstrated no additional tumor

174 Table 2. Prognostic factors in patients with gastrinomas. I. Factors associated with poor prognosis Liver tnetastases - initially or develop with time Extent of liver metastases Lymph node metastases Bone metastases Development of ectopic Cushing's syndrome II. Factors associated with aggressive clinical course of gastrinoma and development of liver metastases Larger primary tumor size (>3 cm) Female gender MEN1 syndrome absent Time from symptom onset to diagnosis is <3 years Markedly increased serum gastrin level Pancreatic primary Flow cytometry results [high S phase (mean, 5.1), low % nontetraploid aneuploid, multiple stem line aneuploid frequent] growth over a mean of 29 mo, 32% had slow growth (1%- 50% volume increase/month) over 19 months and 42% had rapid tumor growth (>50% increase/month). In patients with rapid tumor growth in the liver, 62% died, whereas 0% died in the no- or slow-growth categories [32]. These results of the studies have important implications for treatment. First, liver metastases do not grow rapidly in all patients and in previous studies of antitumor treatment this has not considered as an important variable, yet in the future should be. Second, not all patients with liver metastases from a PET, with the hormones-excess state controlled medically, require vigorous antitumor treatment. Some can be followed without treatment and tumor extent reassessed. Third, patients with PETs with rapidly increasing hepatic metastases, bone metastases or who develop ectopic Cushing's syndrome need vigorous antitumor treatment. Fourth, bone metastases are common in patients with metastatic PETs to the liver and should be routinely sought for by SRS, because their detection almost always alters management. Fifth, PET's resemble carcinoid tumors [33] in that tumor size and primary location are important prognostic factors [29]. Sixth, surgical treatment of localized PET's is justified because the development of liver metastases decreases survival. The last conclusion is supported by a recent study [34] in patients with ZES which demonstrated surgical resection decreased the rate of subsequent development of hepatic metastases. Molecular basis of inherited syndromes causing PET's and the clinical features of the PET's Four different genetic syndromes are reported to be associated with the development of PET's (Table 3). The gene mutated in each of these diseases has now been identified, however, the role of the protein that is altered in these diseases in normal tissues remains unclear (Table 3). Each of these diseases is thought to be caused by loss of function of a tumor suppressor gene [35]. MENIN, the 610 amino acid protein altered in MEN1, is known to occur widely in normal tissues, to be a nucleoprotein; however its function is unclear [36]. von Hippel-Lindau (VHL) disease encodes a 213 amino acid protein that is important in regulating cell growth and differentiation [37]. In von Recklinghausen's disease (neurofibromatosis I) the altered protein, neurofibromin in normal cells functions as a suppressor of the ras signaling cascade [38]. Tuberous sclerosis is caused by mutations in either the 1164 amino acid protein hamartin (TSC1), or the 1807 amino acid protein, tuberin (TSC2) [39]. The COOH terminus of tuberin shows homology to the catalytic N-terminal domains of the GTPase-activating proteins (GAP) which are involved in regulating cell cycle and cell growth [35,39]. Of these 4 different syndromes, the MEN1 syndrome is the most frequently associated with the development of PETs (Tables 1 and 3). In pathologic studies 80%-100% of patients with MEN1 develop nonfunctioning PETs most of which are small (<0.5 cm). In family studies 54% of affected members develop gastrinomas, 21% insulinomas and less than 5% develop the other functional PETs (Table 1). In recent studies PETs are becoming the most common cause of death in patients with MEN1. It is essential that in all patients with a PET, the possibility of MEN 1 be considered, especially in patients with gastrinomas (20%-25% have MEN1), GRFomas (30% with MEN 1), and insulinomas (4%-5% with MEN1) (Table 1). This is essential because the clinical management of the PET, the need for additional studies to detect other endocrinopathies, the natural history, the risk of development of carcinoids which may be malignant (gastric, thymic, bronchial) and the need for screening of other family members, all differ between patients with or without MEN1 [11,40,41]. Mutations have been shown in the MEN1 gene in 75%-94% of patients with clinical findings and a family history suggestive of MEN1; however, the molecular testing involves multiple polymerase chain reactions and is not widely available. Although most patients with MEN1 develop hyperparathyroidism or pituitary disease prior to the development of a symptomatic PET, two recent studies report patients with MEN1 can present only with a PET [42,43]. Therefore all patients with MEN1 should be initially examined for evidence of parathyroid and pituitary hyperfunction, as well as a family history of endocrinopathies sought and these studies should be repeated yearly. If a genetic test for MEN1 becomes widely available it should be performed on all patients with a PET. A controversial point that is still unresolved is the management of gastrinomas in MEN1 patients [11,44]. Whereas there is general agreement that patients with MEN 1 with VIPomas, insulinomas, glucagonomas or GRFomas should undergo resection, if possible, because it is frequently possible to cure these patients; this is not the case with gastrinomas [11,44]. Recent studies demonstrate the gastrinoma in MEN1 patients is present in the duodenum in 70%-80% of patients, in the pancreas in 20%-30%, and it has been suggested these patients may be cured by careful exploration of the duodenum [11]. In a recent prospective study [45] none of the 10 patients with MEN1 and ZES were cured even with routine duodenotomy. This was due to large numbers of duodenal tumors in 43% and because 86% had metastases to lymph nodes. The conclusion from this study was that patients with MEN1 and ZES can not be cured by routine exploration and therefore surgical exploration for cure is not routinely indicated [45]. This conclusion remains controversial [11,44,46]. Patients with von Hippel-Lindau syndrome have PETs in 12%-17% of cases (Table 1) [17,47]. Although occasional insulinomas have been reported, almost all PETs in these patients are nonfunctional PET's. In a recent study [47], 13% of the nonfunctional PETs were malignant and associated

175 Table 3. Genetic syndromes associated with an increased incidence of PETs. Syndrome Multiple endocrine neoplasia-type 1 (MEN1) von Hippel-Lindau disease von Recklinghausen's disease [neurofibromatosis 1 (NF-1)] Tuberous sclerosis Location of gene mutation and gene product 1 Iql3 (encodes 610 amino acid protein, MENIN) 3q25 (encodes 213 amino acid protein) 17ql 1.2 (encodes 2485 amino acid protein, neurofibromin) 9q34(TSCl) (encodes 1164 amino acid protein, hamartin) 16pl3(TSC2) (encodes 1807 amino acid protein, tuberin) PETs seen/frequency 80-100% develop PET: (nonfunctional>gastrinoma>insulinoma) 12-17% develop PETs (almost always nonfunctional) Duodenal somatostatinomas Uncommonly develop PET (nonfunctional and functional) with liver metastases. In this study [47] patients with larger PETs were more likely to have hepatic metastases (P=0.0013). Patients with von Recklinghausen's disease (NF-1) develop duodenal somatostatinomas [5,35]. Recently, this association was reviewed [13] and the findings in 27 patients with duodenal somatostatinomas in NF-1 patients were compared to 29 patients with duodenal somatostatinomas without NF-1 and 32 patients with pancreatic somatostatinomas in the world's literature. Pancreatic somatostatinomas were only rarely associated with NF-1 (6%). Pancreatic somatostatinomas were more frequently associated with liver or lymph node metastases and infrequently had psammona bodies, whereas duodenal tumors in patients without or with NF-1 frequently did [13]. Duodenal somatostatinomas in patients with and without NF-1 were similar in clinical features and histologic changes, however they differed in multihormone production (43% without NF-1 and 18% with NF-1). As discussed earlier, strictly speaking many of these tumors are not somatostatinomas, particularly those in the duodenum, if the presence of a clinical syndrome is required for this diagnosis. Recent insights into the molecular pathogenesis of sporadic PET's (non-inherited) In contrast to a number of common nonendocrine tumors such as carcinomas of the pancreas, breast, stomach, colon and lung, in neither PET's nor carcinoid tumors, have alterations in oncogenes (ras, myc, fos, Jun, Src) or common tumor suppressor genes [p53, retinoblastoma suspectibility gene (Rb)] been found to be generally important in their pathogenesis [35]. Therefore until recently their molecular pathogenesis was almost completely unknown [35]. Recent studies suggest alterations in the MEN1 gene will likely be important in a proportion of sporadic PETs [48-51]. Mutations are reported to occur in 27%-39% of sporadic PETs [49-51 ], and a loss of heterozygosity (LOH) involving the MEN1 gene is reported in 93% of sporadic gastrinomas [52]. At present the molecular basis for this discrepancy is not explained but it has been proposed that hypermethylation of the promoter region, intron-based mutations or mutations in other noncoding areas of the MEN 1 gene may be involved [52]. These data suggest that mutations in the MEN1 gene likely play a critical role in the pathogenesis of at least onethird of sporadic PET's. Recently, inactivation of the pl6/mtsl tumor suppressor gene has been demonstrated in a number of human malignancies, including pancreatic adenocarcinomas [53]. This tumor suppressor gene, located on chromosome 9p21, encodes for a protein that is involved in cell cycle regulation, inhibiting the cell cycle at the G1 -S junction [53]. Mutations, deletions and aberrant methylation of the 5' CpG island of pl6/mtsl are all known to alter the function of this gene and occur in different malignancies [53]. Recently [54] 8 gastrinomas and 4 nonfunctional PETs were studied for alterations in pl6/mtsl. Homozygous deletions were detected in 42% of the PET's, methylation of the 5' CpG island in 58%, but no mutations were identified [54]. Overall 92% of the PETs demonstrated an inactivating mutation in the pl6/mtsl gene, suggesting transcriptional silencing of this gene could be an important event in the molecular pathogenesis of these tumors. References 1. Stridsberg M, Oberg K, Li Q, et al. Measurements of chromogranin A, chromogranin B (secretogranin I), chromogranin C (secretogranin II) and pancreastatin in plasma and urine from patients with carcinoid tumours and endocrine pancreatic tumours. J Endocrinol 1995; 144:49-59. 2. Baudin E, Gigliotti A, Ducreux M, et al. Neuron-specific enolaseand chromogranin A as markers of neuroendocrine tumours. Br J Cancer 1998; 78: 1102-7. 3. Pirker RA, Pont J, Pohnl R, et al. Usefulness of chromogranin A as a marker for detection of relapses of carcinoid tumours. Clin Chem Lab Med 1998; 36: 837-40. 4. Nobels FR, Kwekkeboom DJ, Bouillon R, et al. Chromogranin A: its clinical value as marker of neuroendocrine tumours. Eur J Clin Invest 1998; 28:431-40. 5. Jensen RT, Norton JA. Endocrine tumors of the pancreas. In Gastrointestinal and Liver Disease, Feldman M, Scharschmidt BF, Sleisenger MH, eds. 6th ed. Vol. 1. Philadelphia: W.B. Saunders 1998; 871-94. 6. Mao C, El Attar A, Domenico DR. et al. Carcinoid tumors of the pancreas. Int J Pancreatol 1998; 23: 153-64. 7. Maurer CA, Glaser C, Reubi JC, et al. Carcinoid of the pancreas. Digestion 1997; 58:410-4. 8. Mao C, Carter P, Schaefer P, et al. Malignant islet cell tumor associated with hypercalcemia. Surgery 1995; 117:37-40. 9. Fleury A, Rejou JF, Sauvanet A, et al. Calcitonin-secreting tumors of the pancreas: about six cases. Pancreas 1998; 16:545-50. 10. Rambaud JC, Hautefeuille M, Ruskone A, et al. Diarrhoea due to circulating agents. Clin GI1986; 15:603-29. 11. Jensen RT. Management of the Zollinger-Ellison syndrome in patients with multiple endocrine neoplasia type 1. J Intern Med 1998; 243:477-88. 12. Fraker DL, Jensen RT. Pancreatic endocrine tumors. In Cancer Principles and Practice of Oncology, DeVita VT, Hellman S, Rosenberg SA, eds. 5th ed. Philadelphia: Lippincoa-Raven 1997; 1678-1704. 13. Mao C, Shah A, Hanson DJ, et al. Von Recklinghausen's disease associated with duodenal somatostatinoma: contrast of duodenal versus pancreatic somatostatinomas. J Surg Oncol 1995; 59:67-73. 14. Eriksson B, Oberg K. PPomas and nonfunctioning endocrine pancreatic tumors: clinical presentation, diagnosis, and advances in management. In Endocrine Tumors of the Pancreas: Recent advances in research and management Series: Frontiers in Gastrointestinal Research, Mignon M, Jensen RT, eds. Vol. 23.

176 Basel, Switzerland: S. Karger 1995; 208-22. 15. Krenning EP, Kwekkeboom DJ, Bakker WH, et al. Somatostatin receptor scintigraphy with ["'In-DTPA-D-Phe']- and [ l2j I-Tyr J ]-octreotide: the Rotterdam experience with more than 1000 patients. EurJNuclMed 1993; 20: 716-31. 16. Jensen RT. Peptide therapy. Recent advances in the use of somatostatin and other peptide receptor agonists and antagonists. In Current Clinical Topics in Gastrointestinal Pharmacology, Lewis JH, Dubois A, eds. Maiden, MA: Blackwell Science 1997; 144-223. 17. Gibril F, Reynolds JC, Doppman JL, et al. Somatostatin receptor scintigraphy: its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas: a prospective study. Ann Intern Med 1996; 125: 26-34. 18. van Eijck CH, Lamberts SW, Lemaire LC, et al. The use of somatostatin receptor scintigraphy in the differential diagnosis of pancreatic duct cancers and islet cell tumors. Ann Surg 1996; 224: 119-24. 19. Gibril F, Reynolds JC, Chen CC, et al. Specificity of somatostatin receptor scintigraphy: a prospective study and the effects of false positive localizations on management in patients with gastrinomas. J Nucl Med 1998 (in press). 20. Yu F, Venzon DJ, Serrano J, et al. Prospective study of the clinical course, prognostic factors and survival in patients with longstanding Zollinger-Ellison syndrome. J Clin Oncol 1998 (in press). 21. Metz DC, Strader DB, Orbuch M, et al. Use of omeprazole in Zollinger-Ellison: A prospective nine-year study of efficacy and safety. Aliment Pharmacol Ther 1993; 7: 597-610. 22. Stewart CA, Termanini B, Sutliff VE, et al Assessment of the risk of iron malabsorption in patients with Zollinger-Ellison syndrome treated with long-term gastric acid antisecretory therapy. Aliment Pharmacol Ther 1998; 12: 83-98. 23. Termanini B, Gibril F, Sutliff VE, IH, et al. Effect of long-term gastric acid suppressive therapy on serum vitamin B' 2 levels in patients with Zollinger-Ellison syndrome. Am J Med 1998; 104:422-30. 24. Alexander HR, Fraker DL, Norton JA, et al. Prospective study of somatostatin receptor scintigraphy and its effect on operative outcome in patients with Zollinger-Ellison syndrome. Ann Surg 1998; 228: 228-38. 25. Cadiot G, Lebtahi R, Sarda L, et al. Preoperative detection of duodenal gastrinomas and peripancreatic lymph nodes by somatostatin receptor scintigraphy. Gastroenterology 1996; 111: 845-54. 26. Sugg SL, Norton JA, Fraker DL, et al. A prospective study of intraoperative methods to diagnose and resect duodenal gastrinomas. Ann Surg 1993; 218: 138^4. 27. Doppman JL, Nieman LK, Cutler GB, Jr., et al. Adrenocorticotropic hormone-secreting islet cell tumors: Are they always malignant? Radiology 1994; 190:59-64. 28. Jensen RT, Gardner JD. Gastrinoma. In The Pancreas: Biology, Pathobiology and Disease, Go VLW, DiMagno EP, Gardner JD, et al, eds. 2nd ed. New York: Raven Press 1993; 931-78. 29. Weber HC, Venzon DJ, Lin JT, et al. Determinants of metastatic rate and survival in patients with Zollinger-Ellison syndrome: a prospective long-term study. Gastroenterology 1995; 108: 1637-49. 30. Metz DC, Kuchnio M, Fraker DL, et al. Flow cytometry and Zollinger-Ellison syndrome: relationship to clinical course. Gastroenterology 1993; 105:799-813. 31. Gibril F, Doppman JL, Reynolds JC, et al. Bone metastases in patients with gastrinomas: a prospective study of bone scanning, somatostatin receptor scanning, and MRI in their detection, their frequency, location and effect of their detection on management. J Clin Oncol 1998; 16: 1040-53. 32. Sutliff VE, Doppman JL, Gibril F, et al. Growth of newly diagnosed, untreated metastatic gastrinomas and predictors of growth patterns. J Clin Oncol 1997; 15: 2420-31. 33. Jensen RT, Norton JA. Carcinoid tumors and the carcinoid syndrome. In Cancer Principles and Practice of Oncology, DeVita VT, Jr., Hellman S, Rosenberg SA, eds. Philadelphia: Lippincott-Raven 1997; 1704-23. 34. Fraker DL, Norton JA, Alexander HR, et al. Surgery in Zollinger-Ellison syndrome alters the natural history of gastrinoma. Ann Surg 1994; 220:320-30. 35. Weber HC, Jensen RT. Pancreatic endocrine tumors and carcinoid tumors: recent insights from genetic and molecular biologic studies. In Advances in Pancreatic Disease. Molecular B iology, Diagnosis and Treatment, Dervenis CG, ed. Stuttgart; New York: Georg Thieme Verlag 1996; 55-75. 36. Guru SC, Goldsmith PK, Bums AL, et al. Menin, the product of the MEN1 gene, is a nuclear protein. Proc Nat] Acad Sci USA 1998; 95: 1630-4. 37. Kaelin WG, Jr., Maher ER. The VHL tumour-suppressor gene paradigm. Trends Genet 1998; 14:423-6. 38. Feldkamp M, Gutmann DH, Guha A. Neurofibromatosis type 1: piecing the puzzle together. Can J Neurol Sci 1998; 25:181-91. 39. Young J, Povey S. The genetic basis of tuberous sclerosis. Mol Med Today 1998; 313-9. 40. Metz DC, Jensen RT, Bale AE, et al. Multiple endocrine neoplasia type 1: clinical features and management. In The Parathyroids, Bilezekian JP, Levine MA, Marcus R, eds. New York: Raven Press 1994; 591-646. 41. Norton JA, Jensen RT. Multiple endocrine neoplasia. In Cancer Principles and Practice of Oncology, DeVita VT, Hellman S, Rosenberg SA, eds. 5th ed. Philadelphia: Lippincott-Raven 1997; 1723-9. 42. Benya RV, Metz DC, Venzon DJ, et al. Zollinger-Ellison syndrome can be the initial endocrine manifestation in patients with multiple endocrine neoplasia-type 1. Am J Med 1994; 97:436-44. 43. Cany SE, Helm AK, Arr.ico JA, et al The variable penetrance and spectrum of manifestations of multiple endocrine neoplasia type 1. Surgery 1998; 124: 1106-14. 44. Thompson NW. Current concepts in the surgical management of multiple endocrine neoplasia type 1 pancreatic-duodenal disease. Results in the treatment of 40 patients with Zollinger-Ellison syndrome, hypoglycaemia or both. J Intern Med 1998; 243:495-500. 45. MacFarlane MP, Fraker DL, Alexander HR, et al. A prospective study of surgical resection of duodena] and pancreatic gastrinomas in multiple endocrine neoplasia-type 1. Surgery 1995; 118:973-80. 46. Skogseid B, Oberg K, Akerstrom G, et al. Limited tumor involvement found at multiple endocrine neoplasia type I pancreatic exploration: can it be predicted by preoperative tumor localization? World J Surg 1998; 22:673-8. 47. Libutti SK, Choyke PL, Bartlett DL, et al. Pancreatic neuroendocrine tumors associated with von Hippel-Lindau disease: diagnostic and management recomendations. Surgery 1998; 124: 1153-9. 48. Marx S, Spiegel AM, Skarulis MC, Doppman JL, Collins FS, Liotta LA. Multiple endocrine neoplasia type 1: clinical and genetic topics. Ann Intern Med 1998; 129:484-94. 49. Zhuang ZP, Vortmeyer AO, Pack S, et al. Somatic mutations of the MEN1 tumor suppressor gene detected in sporadic gastrinomas and insulinomas. Cancer Res 1997; 57:4682-6. 50. Hessman O, Lindberg D, Skogseid B, et al. Mutation of the multiple endocrine neoplasia type 1 gene in nonfamilial, malignant tumors of the endocrine pancreas. Cancer Res 1998; 58: 377-9. 51. Wang EH, Ebrahimi SA, Wu AY, et al. Mutation of the MENIN gene in sporadic pancreatic endocrine tumors. Cancer Res 1998; 58:4417-20. 52. Zhuang Z, Vortmeyer AO, Pack S, et al Somatic mutations of the MEN 1 tumor suppressor gene in sporadic gastrinomas and insulinomas. Cancer Res 1997; 57: 4682-6. 53. Ruas M, Peters G. The p 16INK4a/CDKN2A tumor suppressor and its relatives. Biochim Biophys Acta 1998; 1378: Fl 15-77. 54. Muscarella P, Melvin WS, Fisher WE, et al. Genetic alterations in gastrinomas and nonfunctioning pancreatic neuroendocrine tumors: an analysis of pi 6/MTS1 tumor suppressor gene inactivation. Cancer Res 1998; 58:237-40. Correspondence to: Robert T. Jensen, M.D. Digestive Diseases Branch, NIDDK National Institutes of Health Bldg. 10,Rm.9C-103 10 CENTER DRMSC 1804 BETHESDA MD 20892-1804 USA E-mail: robertj@bdglo.niddk.nih.gov