- Email: [email protected]
Oncogenes, growth factors, receptor expression and proliferation markers in digestive neuroendocrine tumours. A critical reappraisal
Annals of Oncology 12 (Suppl. 2): S13-S17, 2001. O 2001 Kluwer Academic Publishers. Printed in the Netherlands.
Review Oncogenes, growth factors, receptor expression and proliferation markers in digestive neuroendocrine tumours. A critical reappraisal G. Delle Fave & V. D. Corleto Department of Gastrointestinal and Liver Diseases, II School of Medicine, 'La Sapienza' University, Rome, Italy
dNETs. The role of various growth factors, novel oncogenes and tumour suppressor genes have also been investigated. Background: The main characteristic of the digestive neuro- However, results from these studies are non-conclusive and to endocrine tumours (dNETs) is the low proliferating activity, date the molecular pathogenesis of these tumours has not been clarified. Studies on somatostatin receptor expression and even in the presence of malignant, metastatic behavior. Patients and methods: Considering that dNETs are rare synthetic analogues, as growth inhibitors in dNETs, although diseases, relatively numerous studies, often including a con- promising, have not reproduced in vivo all the antiproliferative spicuous number of patients, have recently investigated the effects showed in in vitro models. molecular mechanisms of neuroendocrine tumour genesis. Conclusion: Although various functional genes and molecResults: In contrast to non-endocrine tumours of the diges- ular mechanisms have been investigated in dNETs, to date the tive system such as carcinoma of the pancreas, colon and molecular pathogenesis of these tumours remains to be elucistomach, dNETs do not show alterations in oncogenes (ras, dated. Myc, fos, jun, Src) or in common tumor suppressor genes [p53, retinoblastoma suspectibility gene (Rb)]. MEN-1 gene altera- Key words: digestive neuroendocrine tumour, growth factor, tions will likely be important in a proportion of sporadic oncogene, tumour suppressor gene Summary
Although the vast majority of dNET tumours are represented by well- differentiated cells, with the low rate of proliferating cells they often present metastases at the time of the diagnosis. This is one of the most intriguing characteristics of dNET tumours and has triggered the scientific interest of many groups who aim to demonstrate specific molecular features that can explain which mechanisms can account for the ability of these tumour cells to detach from primary malignancy and gain access to the surrounding structures and liver. However, except for tumours developing in patients with specific genomic gene alterations, such as MEN-1 syndrome and von Hippel Lindau disease (VHL), the molecular events determining the uncontrolled growth of neuroendocrine cells are still unclear.
Digestive neuroendocrine tumours of the gastrointestinal tract (dNETs) are relatively rare tumours and due to the heterogeneity of their clinical and biological aspects their study, although fascinating, is complex. The pancreas and the mucosa of the gastrointestinal tract contain as many as 16 cell types that belong to the so-called diffuse endocrine system (DES), characterized by the production of peptides or amines from which in theory such tumours can develop. However, in practice, tumour transformation has been demonstrated only for some of the DES cells. For diagnostic purposes most of these tumours are sufficiently characterized by histological features showing solid, trabecular or glandular arrangement of well-differentiated cells. For these characteristics specific immunohistochemical stains for general neuroendocrine markers (chromogranins, synaptophy- Tumour suppressor genes and oncogenes sin, or neuron-specific enolase) and hormonal products, are usually able to identify the neuroendocrine origin of The majority of neuroendocrine tumours are sporadic the tumours. They also provide the basis of the so-called and to date no clear hypothesis on tumour genesis and/ 'morphofunctional' classification that has been recently or tumour progression has been made. revised [1]. In this classification tumours are divided by As in most common carcinomas, a number of tumour site of origin (stomach, pancreas, etc.) and classified by suppression genes and oncogenes have been investigated their biological behaviour; furthermore the term 'carci- in dNETs. However, due to the great variety of these noid' has been abandoned in favour of tumour or carci- tumours, their relative rarity and the differences in noma, and anatomic, clinical and functional data are tissues sampled analysed, no definitive conclusion can utilized in combination. be made.
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Introduction
14
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Mutations within the p53 gene are known to be due to considered to act as tumour suppressor genes, located common genetic alteration, occurring in about half of on chromosomes Ilql3 and 3p25, respectively. After its all types of cancers arising from a wide spectrum of cloning (1997) MEN-1 gene has been investigated in tissues. In dNETs the p53 somatic mutation and/or different types of sporadic neuroendocrine tumours and over-expression have been reported mostly in the late its mutation has been recently reported in 16%—42% of stage, furthermore, when present, its over-expression patients [10]. In general, when studies involve a small has no relationship with the clinical-pathological fea- number of patients it is likely that the proportion of gene tures [2]. At the present we can assume that p53 muta- alterations has been over emphasised, whereas when tion/over-expression may be unimportant in the genesis studies are performed on a relative large number of of dNETs [2, 3]. Although the function of the K-ras patients, the percentage of alteration of MEN-1 gene is pathway remains incompletely understood, over 90% of better defined and conclusions on its significance in the pancreatic non-endocrine cancers show mutations of pathogenesis of sporadic neuroendocrine tumours can this oncogene. In dNETs K-ras mutation has been spor- be made. Specifically, MEN-1 gene was mutated in 31% adically reported [4, 5] and does not seem to be involved of 51 sporadic gastrinomas recently investigated, in the in these tumours. Other common oncogenes such as same study the presence of a MEN-1 gene mutation myc,fas,jun, bcl-2 and src, in contrast to their involve- correlated only with primary tumour location and not ment in the tumor genesis of non-endocrine tumours, with clinical characteristics and /or tumour growth patare generally not important in the molecular pathogen- tern [11]. Results from studies on von Hippel-Lindau esis of dNETs. Hypothetically, their deregulation may syndrome are interesting - not for the frequency of represent one of the pathogenic events in the malignant pancreatic endocrine tumours in VHL patients, nor for transformation and/or progression of only few atypical VHL gene alterations found in a sporadic form of forms of these tumours [6]. The cyclin-dependent kinases dNETs, in fact both are very low - but because from are enzymes that regulate various faces of the cell these studies the presence of a different new tumour division cycle, and the role of their inhibitors such as suppressor gene, other than VHL on chromosome 3p, pl5, pl6, p27 and p21 as tumour suppressor genes have has been proposed. Data supporting these conclusions recently also been evaluated in dNETs. A high frequency come from studies on non-functioning pancreatic tuof gene deletion and/or mutation of gene coding for mours [12] in which the contemporary presence of rare these inhibitors has important implications in human or absent mutations on VHL gene were accompanied by cancer genesis, and in this regard they can be function- a 30% allelic loss on chromosome 3p. Then considering ally considered as tumour suppressor genes. In partic- that deletion in one allele is accompanied by mutation in ular, high p27 and low p21 expression has been reported the other, the existence of a different mutated gene has in differentiated dNETs, suggesting a potential role of been suggested in these cases. p27 as inhibitor of cell proliferation in these type of pi6, a tumor suppressor gene located on chromosome tumours [7]. During the last few years, the putative 9p21, encodes for an inhibitor of cyclin-dependent kintumour suppressor gene, DPC4, successively designed ase4 that, when activated, inhibits cell progression at the as Smad4 and involved in signal transduction from the Gi S junction by phosphorylating protein expression of TGF-beta family of cytokines, has been demonstrated the retinoblastoma gene (Rb gene). Recent studies in to be deleted and/or inactivated in up 50% of pancreatic non-endocrine tumours reported that in some tumours exocrine carcinomas. Due to the anatomical relationpl6 gene alteration and/or loss of pl6 protein can affect ship existing between exocrine and endocrine pancreas, prognosis, recurrence, tumour aggressiveness and paSmad4 gene has been investigated also in pancreatic tient's survival [13]. pl6 inactivation has also been studied endocrine tumours. One study reported Smad4 alterations (deletions and/or mutations) in five out of nine in a number of neuroendocrine tumours. Specifically, non-functioning pancreatic tumours and none in eleven homozygous deletions and/or hypermethylation have insulinomas, three gastrinomas and two VIPomas [8]. been reported in 92% of a small number of gastrinomas A more recent study tested, among other pancreatic and nonfunctional tumours [14]. In a successive study, exocrine tumours, 41 cases of non-functioning endo- performed on a group of 44 gastrinomas, only hypercrine tumours and found no Smad4 gene alteration in methylation of a pi 6 gene promoter island was found these tumours whatsoever [9]. Even if it is difficult to in 52% of the tumours [15]. However, in another study explain these different results, it is likely that different on 41 pancreatic endocrine tumours (30 non-functioning, 8 insulinomas, 1 gastrinoma and 1 Vipoma) pl6 methods could account for such discrepancy. mutation was found only in 1 insulinoma and no gene Causal genes that are altered in familial cancer syn- alteration nor hypermethylation or homozygous deledromes are usually altered in sporadic disease counter- tion was shown in any of the other tumours [9]. Then, parts. On this basis, knowing that patients with familial although to date pi6 gene hypermethylation is the most germ-line alteration in MEN-1 and VHL genes often common gene alteration reported in gastrinomas, it develop gastrointestinal and/or pancreatic endocrine seems that this alteration is, for unknown reasons, tumours (Table 1), it may be expected that these genes limited mostly to functional neuroendocrine tumours. could be modified in the respective sporadic form of endocrine neoplasms. Both MEN-1 and VHL genes are
15 Table I. Gastrointestinal endocrine cell types, their localization and relation to hormonal syndromes. Tumour type
Cell
Main prod-
type
uct
Main localization Pancreas
Welldifferentiated
Stomach Corpus/ fundus
Small intestine Antrum
Duod
Jejun
Large intestine Ileum
App
Colon
Rectum
B
Insulin
+
Insulinoma (PHH)
A
Glucagon
+
Glucagonoma
+
PP:oma
PP
PP
Neuron
VIP
D
Somatostatin
X
Unknown
+
P
Unknown
+
CCK
CCK
VIP:oma (pancreas) +
+
+
+
+
+
Somatostatinoma -
+
-
+ +
+
+
+
+
+
-
S
Secretin
EC
Serotonin
ECL
Histamin
G
Gastrin
GIP
GIP
+
+
-
M
Motilin
+
+
-
N
Neurotensin
+
+
L
Enteroglucagon/PYY
+
+
+
+
+
+
+
+
+
+
s/i
+
+
+
+
+
+
+
+
ECL cell carcinoid (atypical syndr) +
+
+
EC cell carcinoid (typical syndr)
+
Gastrinoma (ZES)
+
_
Abbreviations: s/i - small-intermediate-sized cells; EC - enterochromaffin; ECL - enterochromaffin-like cell; PP - pancreatic polypeptide; VIP vasoactive intestinal polypeptide; CCK - cholecystokinin; GIP - gastric inhibitory polypeptide; PYY - PP-like peptide with N-terminal tyrosine amide; PHH - persistent hyperinsuhnemic hypoglycemia; ZES - Zollinger-Ellison syndrome. Modified after Rindi et al (1999) [6],
Growth factors
Growth factors have been variously investigated in different types of dNETs, specifically, production of such factors as alpha and beta fibroblast growth factor (FGF) and transforming growth factor (TGF), plateletderived growth factor (PDGF) and insulin-like growth factor-I (IGF-I) have been demonstrated by most of dNETs investigated [16, 17]. Although involved in the autocrine stimulation of tumour cells, a specific function for these factors, either in the genesis and/or in the progression of dNETs are not yet elucidated. Interestingly, a high expression of TGF alpha in a group of gastro-intestinal carcinoids has been reported. However, this expression was not accompanied by an intact specific receptor molecule i.e. epidermal growth factor receptor (EGFR), making ineffective the TGF growth factor function and suggesting that this may be related to their indolent behaviour. Among the various growth factors promoting angiogenesis, vascular endothelial growth factor (VEGF), a specific endothelial cell mito-
gen agent strictly correlated to the tumour progression, has been studied in a group of 48 neuroendocrine tumours. Its expression (from 25% to > 50% of positive cells) was found in 78% of 28 gastrointestinal carcinoids; specifically, all the midgut carcinoids were intensely positive, whereas it was found in only 25% of either functioning and non-functioning pancreatic neuroendocrine tumours [18]. These results suggest that VEGF in association with the multiple growth factors synthesized by these tumours may play an important, although not yet understood, role in tumour angiogenesis and indirectly in tumour growth, especially in midgut carcinoids. Furthermore, in nude mice, xenotransplanted with human intestinal endocrine tumour, a protective effect of specific VEGF-antibody on tumour progression has been reported. In fact only in 2 out of 19 VEGFantibody treated mice developed liver metastasis compared to almost all untreated mice [19].
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Poordifferentiated
Possible syndrome
16 Peptide receptors
Conclusions The molecular pathogenesis of neuroendocrine tumours of the gastrointestinal tract, to date remains largely unknown. Molecular mechanisms involved in the tumour genesis of neuroendocrine compared to non-endocrine tumours such as colo-rectal and pancreatic cancer are different. However, the significant recent advances, involving localization methods, natural history and new treatments, have prompted the attention of many studies on the peculiar characteristics of these tumours that could improve the over-all knowledge on dNETs.
Acknowledgements This study was supported by an inter-university cofinanced program 9906218982-005 (1999) from the Italian Ministry for University and Scientific and Technological Research (MURST), by a grant for finalised research from the Italian Ministry for Health 2000 and Fondazione Italiana per le Malattie Digestive (FIMAD). References 1. Solcia E, Kloppel G, Sobin LH. Histological Typing of Endocrine Tumours. Berlin/Heidelberg: Springer-Verlag 2000. 2. Wang D-G, Johnston CF, Anderson N et al. Overexpression of the tumor suppressor gene p53 is not implicated in neuroendocrine tumour carcinogenesis. J Pathol 1995; 175: 397-401. 3. Bartz C, Ziske C, Wiedenmann B et al. p53 tumour suppressor gene expression in pancreatic neuroendocrine tumour cells. Gut 1996; 38: 403-9. 4. Beghelli S, Pelosi G, Zamboni G et al. Pancreatic endocrine tumours: Evidence for a tumour suppressor pathogenesis and for a tumour suppressor gene on chromosome 17p. J Pathol 1998; 186: 41-50. 5. Ebert MP, Hoffmann J, Scheider-Stock R et al. Analysis of K-ras gene mutations in rare pancreatic and ampullary tumours. Eur J Gastroenterol Hepatol 1998; 10: 1025-9.
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Increasing evidence exists that growth and proliferations, as well as the functional properties of various tumour cells, are modulated by gastrointestinal peptides. The low availability of stable cultured neuroendocrine tumour cell lines still represent a limiting factor in the study of the specific function of these receptors expressed on neuroendocrine tumours. However, for vasoactive intestinal peptide receptors (VIP-1 and VIP-2) and most of all for somatostatin receptors (ssts), numerous studies attempting to elucidate basic mechanisms of receptor functions and clinical applications of specific peptide agonists have been produced during the last few years. VIP receptors have been demonstrated, mostly by [123I]-VIP receptor scintigraphy, to be overexpressed on tumour cells of many intestinal adenocarcinomas as colorectal, pancreatic and gastric cancers, and in most of primary and metastatic digestive neuroendocrine tumours [20]. Although some studies suggest that VIP can regulate the growth and function of tumour cells [21], to date, no specific synthetic agonists for VIP-1 and/or VIP-2 receptor subtypes are available for clinical use. Therefore, radio-iodinated VIP for scintigraphic tumour visualization is the only clinical application available until now. On the contrary, during these years, studies on somatostatin and its receptor family are literally exploding. Many are the reasons accounting for the great interest that somatostatin has stimulated among basic and clinical scientists. Firstly: somatostatin mostly acts as an inhibitory factor that regulates a large number of physiological functions, such as inhibition of endocrine and exocrine secretions, modulation of neurotransmission, motor and cognitive function, inhibition of intestinal motility and intestinal smooth muscle cells contractility [22, 23], absorption of nutrients and ions, vascular contractility and cell proliferations [24, 25]. Secondly: during the last few years five different somatostatin receptor subtypes have been cloned and these are usually expressed in a cell tissue specific manner. Thirdly: synthetic somatostatin analogues, with relative subtype receptor specificity are available either for diagnostic (scintigraphy) and/or for therapeutic purposes in many diseases. Neuroendocrine tumours have been demonstrated to possess a high number of somatostatin receptors i.e., 100 times more than normal tissues. This property makes them a target of somatostatin agonists in the attempt to inhibit both secretion and growth by these tumour cells. In human digestive neuroendocrine tumour cells somatostatin and octreotide inhibit neurotransmitters and hormone secretions through inhibition of cellular calcium channels via sst2 and sst5 subtypes activation [26]. The antiproliferative action of somatostatin and its analogues (octreotide, lanreotide) induce Gi cell cycle arrest, activating distinct signal transduction pathways depending on receptor subtype. Sst2 antiproliferative effects are mediated by activation of tyrosine phosphatase SHP-1 and dephosphorylation of activated growth factor receptors (27); sstj mediates cell
growth arrest by stimulation of tyrosine phosphatase SHP-2, activation of MAP kinese pathway and induction of the cyclin-dependent kinase inhibitor p21 [28]; sst5 acts by a mechanism involving a dephosphorylation cascade, inhibition of cGMP-dependent protein kinase G and MAP kinase [29]. The antiproliferative effect of somatostatin can also result from apoptosis induced by sst3 subtype [30] or by indirect effects of the peptide resulting from the inhibition of secretion of growth promoting hormones and growth factors which specifically regulate tumour growth. These effects, although very impressive, have been mostly evaluated in experimental 'in vitro' conditions in somatostatin receptor subtypes transfected cells. However, more recent data on living patients with dNETs analysed prospectively indicate that the most beneficial antiproliferative effect of somatostatin analogues treatment account for 36% to 70% of tumour growth stabilization lasting from months to years [31, 32].
17 21. Pincus DW, DiCicco-Bloom EM, Blank IB et al. Vasoactive intestinal polypeptide regulates mitosis, differentiation and survival of cultured sympathetic neuroblasts. Nature 1990; 343: 56418. 22. Gu ZF, Corleto VD, Mantey SA et al. Somatostatin receptor subtype 3 mediates the inhibitory action of somatostatin on gastric smooth muscle cells. Am J Physiol 1995; 268: G739-G748. 23. Corleto VD, Severi C, Coy DH, et al. Colonic smooth muscle cells possess a different subtype of somatostatin receptor than gastric smooth muscle cells. Am J Physiol 1997; 272: G689-G697. 24. Patel YC. Molecular pharmacology of somatostatin receptor subtypes. J Endocrinol Invest 1997; 20: 348-67. 25. Benali N, Ferjoux G, Puente E et al. Somatostatin receptors. Digestion 2000; 62 (Suppl 1): 27-32. 26. Glassmeier G, Hopfner M, Riecken EO et al. Inhibition of L-type calcium channels by octreotide in isolated human neuroendocrine tumor cells of the gut. Biochem Biophys Res Commun 1998; 250: 511-5. 27. Bousquet C, Delesque N, Lopez F, et al. sst2 somatostatin receptor mediates negative regulation of insulin receptor signaling through the tyrosin phosphatases Shp-1. J Biol Chem 1998; 273: 7099-106. 28. FlorioT, Yao H, Carey KD. Somatostatin activation of mitogenactivated protein kinase via somatostatin receptorl (sstO. Mol Endocrinol 1999; 13: 24-37. 29. Cordelier P, Esteve JP, Bousquet C et al. Characterization of the antiproliferative signal mediated by the somatostatin receptor subtype 5. Proc Natl Acad Sci USA 1997; 94: 9343-8. 30. Sharma K, Srikant CB. Induction of wild-type p53, Bax, and acidic endonuclease durino somatostatin-signaled apoptosis in MCF-7 human breast cancer cells. Int J Cancer 1998; 76: 259-66. 31. Corleto VD, Angeletti S, Schillaci O, et al. Long-term octreotide treatment of metastatic carcinoid tumour. Ann Oncol 2000; 11: 491-3. 32. Arnold R, Simon B, Wied M. Treatment of neuroendocrine GEP tumours with somatostatin analogues. Digestion 2000; 62 (Suppl 1): 84-91.
Correspondence to: Gianfranco Delle Fave, MD Gastroenterology Department 11° Clinica Medica, Policlinico Umberto 1° 00161 Viale del Policlinico 155 Rome Italy E-mail: [email protected]
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
6. Weber HC, Jensen RT. Pancreatic endocrine tumors and carcinoid tumors: recent insights from genetic and molecular biologic studies. In Dervenis CG (eds): Advances in Pancreatic Disease. Molecular Biology, Diagnosis and Treatment Stuttgart/New York: GeorgThiemeVerlag 1996; 55-75. 7. Canavese G, Azzoni C, Pizzi S et al. p27: A potential main inhibitor of cell proliferation in digestive endocrine tumors but not a marker of benign behavior. Hum Pathol 2000 (in press). 8. Bartsch D, Hahn SA, Danichevski KD et al. Mutation of the DPC4/Smad4 gene in neuroendocrine pancreatic tumors. Oncogenesl999; 18:2367-71. 9. Moore PS, Orlandini S, Zamboni G et al. Pancreatic tumours: Molecular pathways implicated in ductal cancer are involved in ampullary but not in esocrine nonductular or endocrine tumourigenesis. Cancer Res 2000 (in press). 10. Wang EH, Ebrahimi SA, Wu AYet al. Mutation of the MENIN gene in sporadic pancreatic endocrine tumors. Cancer Res 1998; 58: 4417-20. 11. Goebel SU, Heppner C, Burns AL et al. Genotype/phenotype correlation of multiple endocrine neoplasia type 1 gene mutations in sporadic gastrinomas. J Clin Endocrinol Metab 2000; 85: 11623. 12. Moore PS, Missiaglia E, Antonello D et al. The role of disease causing genes in sporadic pancreatic endocrine tumors. MEN1 and VHL. J Clin Endocrinol Metab 2000 (submitted). 13. Ruas M, Peters G. The pl61NK4a/CDKN2a tumor suppressor and its relatives. Biochim Biophys Acta 1998; 1378: F115-F117. 14. Muscarella P, Melvin WS, Fisher WE et al. Genetic alterations in gastrinomas and non-functioning pancreatic neuroendocrine tumors: An analysis of pl6/MTSl tumor suppressor gene inactivation. Cancer Res 1998; 58: 237^»0. 15. Serrano J, Goebel SU, Peghini PL et al. Alteration in the pl6 INK4a/CDK.N2A tumor suppressor gene in gastrinomas. J Clin Endocrinol Metab 2000; 85: 4146-56. 16. Wulbrand U, Wied M, Zofel P et al. Growth factor receptor expression in human gastroenteropancreatic neuroendocrine tumours. Eur J Clin Invest 1998; 28: 1038-^9. 17. Oberg K. Expression of growth factors and their receptors in neuroendocrine gut and pancreatic tumors, and prognostic factors for survival. Ann NYAcad Sd 1994; 733: 46-55. 18. Terris B, Scoazec JY, Rubbia L at al. Expression of vascular endothelial growth factor in digestive neuroendocrine tumours. Histopathology 1998; 32: 133-8. 19. Konno H, Arai T, TanakaTet al. Antitumor effect of neutralizing antibody to vascular endothelial growth factor on liver metastasis of endocrine neoplasm. Jpn J Cancer Res 1998; 89: 939-9. 20. Virgolmi I. Receptor nuclear medicine: Vasoactive peptide and somatostatin receptor scintigraphy for diagnosis and treatment of tumor patients. Eur J Clin Invest 1997; 27: 793-800.
Review Oncogenes, growth factors, receptor expression and proliferation markers in digestive neuroendocrine tumours. A critical reappraisal G. Delle Fave & V. D. Corleto Department of Gastrointestinal and Liver Diseases, II School of Medicine, 'La Sapienza' University, Rome, Italy
dNETs. The role of various growth factors, novel oncogenes and tumour suppressor genes have also been investigated. Background: The main characteristic of the digestive neuro- However, results from these studies are non-conclusive and to endocrine tumours (dNETs) is the low proliferating activity, date the molecular pathogenesis of these tumours has not been clarified. Studies on somatostatin receptor expression and even in the presence of malignant, metastatic behavior. Patients and methods: Considering that dNETs are rare synthetic analogues, as growth inhibitors in dNETs, although diseases, relatively numerous studies, often including a con- promising, have not reproduced in vivo all the antiproliferative spicuous number of patients, have recently investigated the effects showed in in vitro models. molecular mechanisms of neuroendocrine tumour genesis. Conclusion: Although various functional genes and molecResults: In contrast to non-endocrine tumours of the diges- ular mechanisms have been investigated in dNETs, to date the tive system such as carcinoma of the pancreas, colon and molecular pathogenesis of these tumours remains to be elucistomach, dNETs do not show alterations in oncogenes (ras, dated. Myc, fos, jun, Src) or in common tumor suppressor genes [p53, retinoblastoma suspectibility gene (Rb)]. MEN-1 gene altera- Key words: digestive neuroendocrine tumour, growth factor, tions will likely be important in a proportion of sporadic oncogene, tumour suppressor gene Summary
Although the vast majority of dNET tumours are represented by well- differentiated cells, with the low rate of proliferating cells they often present metastases at the time of the diagnosis. This is one of the most intriguing characteristics of dNET tumours and has triggered the scientific interest of many groups who aim to demonstrate specific molecular features that can explain which mechanisms can account for the ability of these tumour cells to detach from primary malignancy and gain access to the surrounding structures and liver. However, except for tumours developing in patients with specific genomic gene alterations, such as MEN-1 syndrome and von Hippel Lindau disease (VHL), the molecular events determining the uncontrolled growth of neuroendocrine cells are still unclear.
Digestive neuroendocrine tumours of the gastrointestinal tract (dNETs) are relatively rare tumours and due to the heterogeneity of their clinical and biological aspects their study, although fascinating, is complex. The pancreas and the mucosa of the gastrointestinal tract contain as many as 16 cell types that belong to the so-called diffuse endocrine system (DES), characterized by the production of peptides or amines from which in theory such tumours can develop. However, in practice, tumour transformation has been demonstrated only for some of the DES cells. For diagnostic purposes most of these tumours are sufficiently characterized by histological features showing solid, trabecular or glandular arrangement of well-differentiated cells. For these characteristics specific immunohistochemical stains for general neuroendocrine markers (chromogranins, synaptophy- Tumour suppressor genes and oncogenes sin, or neuron-specific enolase) and hormonal products, are usually able to identify the neuroendocrine origin of The majority of neuroendocrine tumours are sporadic the tumours. They also provide the basis of the so-called and to date no clear hypothesis on tumour genesis and/ 'morphofunctional' classification that has been recently or tumour progression has been made. revised [1]. In this classification tumours are divided by As in most common carcinomas, a number of tumour site of origin (stomach, pancreas, etc.) and classified by suppression genes and oncogenes have been investigated their biological behaviour; furthermore the term 'carci- in dNETs. However, due to the great variety of these noid' has been abandoned in favour of tumour or carci- tumours, their relative rarity and the differences in noma, and anatomic, clinical and functional data are tissues sampled analysed, no definitive conclusion can utilized in combination. be made.
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Introduction
14
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Mutations within the p53 gene are known to be due to considered to act as tumour suppressor genes, located common genetic alteration, occurring in about half of on chromosomes Ilql3 and 3p25, respectively. After its all types of cancers arising from a wide spectrum of cloning (1997) MEN-1 gene has been investigated in tissues. In dNETs the p53 somatic mutation and/or different types of sporadic neuroendocrine tumours and over-expression have been reported mostly in the late its mutation has been recently reported in 16%—42% of stage, furthermore, when present, its over-expression patients [10]. In general, when studies involve a small has no relationship with the clinical-pathological fea- number of patients it is likely that the proportion of gene tures [2]. At the present we can assume that p53 muta- alterations has been over emphasised, whereas when tion/over-expression may be unimportant in the genesis studies are performed on a relative large number of of dNETs [2, 3]. Although the function of the K-ras patients, the percentage of alteration of MEN-1 gene is pathway remains incompletely understood, over 90% of better defined and conclusions on its significance in the pancreatic non-endocrine cancers show mutations of pathogenesis of sporadic neuroendocrine tumours can this oncogene. In dNETs K-ras mutation has been spor- be made. Specifically, MEN-1 gene was mutated in 31% adically reported [4, 5] and does not seem to be involved of 51 sporadic gastrinomas recently investigated, in the in these tumours. Other common oncogenes such as same study the presence of a MEN-1 gene mutation myc,fas,jun, bcl-2 and src, in contrast to their involve- correlated only with primary tumour location and not ment in the tumor genesis of non-endocrine tumours, with clinical characteristics and /or tumour growth patare generally not important in the molecular pathogen- tern [11]. Results from studies on von Hippel-Lindau esis of dNETs. Hypothetically, their deregulation may syndrome are interesting - not for the frequency of represent one of the pathogenic events in the malignant pancreatic endocrine tumours in VHL patients, nor for transformation and/or progression of only few atypical VHL gene alterations found in a sporadic form of forms of these tumours [6]. The cyclin-dependent kinases dNETs, in fact both are very low - but because from are enzymes that regulate various faces of the cell these studies the presence of a different new tumour division cycle, and the role of their inhibitors such as suppressor gene, other than VHL on chromosome 3p, pl5, pl6, p27 and p21 as tumour suppressor genes have has been proposed. Data supporting these conclusions recently also been evaluated in dNETs. A high frequency come from studies on non-functioning pancreatic tuof gene deletion and/or mutation of gene coding for mours [12] in which the contemporary presence of rare these inhibitors has important implications in human or absent mutations on VHL gene were accompanied by cancer genesis, and in this regard they can be function- a 30% allelic loss on chromosome 3p. Then considering ally considered as tumour suppressor genes. In partic- that deletion in one allele is accompanied by mutation in ular, high p27 and low p21 expression has been reported the other, the existence of a different mutated gene has in differentiated dNETs, suggesting a potential role of been suggested in these cases. p27 as inhibitor of cell proliferation in these type of pi6, a tumor suppressor gene located on chromosome tumours [7]. During the last few years, the putative 9p21, encodes for an inhibitor of cyclin-dependent kintumour suppressor gene, DPC4, successively designed ase4 that, when activated, inhibits cell progression at the as Smad4 and involved in signal transduction from the Gi S junction by phosphorylating protein expression of TGF-beta family of cytokines, has been demonstrated the retinoblastoma gene (Rb gene). Recent studies in to be deleted and/or inactivated in up 50% of pancreatic non-endocrine tumours reported that in some tumours exocrine carcinomas. Due to the anatomical relationpl6 gene alteration and/or loss of pl6 protein can affect ship existing between exocrine and endocrine pancreas, prognosis, recurrence, tumour aggressiveness and paSmad4 gene has been investigated also in pancreatic tient's survival [13]. pl6 inactivation has also been studied endocrine tumours. One study reported Smad4 alterations (deletions and/or mutations) in five out of nine in a number of neuroendocrine tumours. Specifically, non-functioning pancreatic tumours and none in eleven homozygous deletions and/or hypermethylation have insulinomas, three gastrinomas and two VIPomas [8]. been reported in 92% of a small number of gastrinomas A more recent study tested, among other pancreatic and nonfunctional tumours [14]. In a successive study, exocrine tumours, 41 cases of non-functioning endo- performed on a group of 44 gastrinomas, only hypercrine tumours and found no Smad4 gene alteration in methylation of a pi 6 gene promoter island was found these tumours whatsoever [9]. Even if it is difficult to in 52% of the tumours [15]. However, in another study explain these different results, it is likely that different on 41 pancreatic endocrine tumours (30 non-functioning, 8 insulinomas, 1 gastrinoma and 1 Vipoma) pl6 methods could account for such discrepancy. mutation was found only in 1 insulinoma and no gene Causal genes that are altered in familial cancer syn- alteration nor hypermethylation or homozygous deledromes are usually altered in sporadic disease counter- tion was shown in any of the other tumours [9]. Then, parts. On this basis, knowing that patients with familial although to date pi6 gene hypermethylation is the most germ-line alteration in MEN-1 and VHL genes often common gene alteration reported in gastrinomas, it develop gastrointestinal and/or pancreatic endocrine seems that this alteration is, for unknown reasons, tumours (Table 1), it may be expected that these genes limited mostly to functional neuroendocrine tumours. could be modified in the respective sporadic form of endocrine neoplasms. Both MEN-1 and VHL genes are
15 Table I. Gastrointestinal endocrine cell types, their localization and relation to hormonal syndromes. Tumour type
Cell
Main prod-
type
uct
Main localization Pancreas
Welldifferentiated
Stomach Corpus/ fundus
Small intestine Antrum
Duod
Jejun
Large intestine Ileum
App
Colon
Rectum
B
Insulin
+
Insulinoma (PHH)
A
Glucagon
+
Glucagonoma
+
PP:oma
PP
PP
Neuron
VIP
D
Somatostatin
X
Unknown
+
P
Unknown
+
CCK
CCK
VIP:oma (pancreas) +
+
+
+
+
+
Somatostatinoma -
+
-
+ +
+
+
+
+
+
-
S
Secretin
EC
Serotonin
ECL
Histamin
G
Gastrin
GIP
GIP
+
+
-
M
Motilin
+
+
-
N
Neurotensin
+
+
L
Enteroglucagon/PYY
+
+
+
+
+
+
+
+
+
+
s/i
+
+
+
+
+
+
+
+
ECL cell carcinoid (atypical syndr) +
+
+
EC cell carcinoid (typical syndr)
+
Gastrinoma (ZES)
+
_
Abbreviations: s/i - small-intermediate-sized cells; EC - enterochromaffin; ECL - enterochromaffin-like cell; PP - pancreatic polypeptide; VIP vasoactive intestinal polypeptide; CCK - cholecystokinin; GIP - gastric inhibitory polypeptide; PYY - PP-like peptide with N-terminal tyrosine amide; PHH - persistent hyperinsuhnemic hypoglycemia; ZES - Zollinger-Ellison syndrome. Modified after Rindi et al (1999) [6],
Growth factors
Growth factors have been variously investigated in different types of dNETs, specifically, production of such factors as alpha and beta fibroblast growth factor (FGF) and transforming growth factor (TGF), plateletderived growth factor (PDGF) and insulin-like growth factor-I (IGF-I) have been demonstrated by most of dNETs investigated [16, 17]. Although involved in the autocrine stimulation of tumour cells, a specific function for these factors, either in the genesis and/or in the progression of dNETs are not yet elucidated. Interestingly, a high expression of TGF alpha in a group of gastro-intestinal carcinoids has been reported. However, this expression was not accompanied by an intact specific receptor molecule i.e. epidermal growth factor receptor (EGFR), making ineffective the TGF growth factor function and suggesting that this may be related to their indolent behaviour. Among the various growth factors promoting angiogenesis, vascular endothelial growth factor (VEGF), a specific endothelial cell mito-
gen agent strictly correlated to the tumour progression, has been studied in a group of 48 neuroendocrine tumours. Its expression (from 25% to > 50% of positive cells) was found in 78% of 28 gastrointestinal carcinoids; specifically, all the midgut carcinoids were intensely positive, whereas it was found in only 25% of either functioning and non-functioning pancreatic neuroendocrine tumours [18]. These results suggest that VEGF in association with the multiple growth factors synthesized by these tumours may play an important, although not yet understood, role in tumour angiogenesis and indirectly in tumour growth, especially in midgut carcinoids. Furthermore, in nude mice, xenotransplanted with human intestinal endocrine tumour, a protective effect of specific VEGF-antibody on tumour progression has been reported. In fact only in 2 out of 19 VEGFantibody treated mice developed liver metastasis compared to almost all untreated mice [19].
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Poordifferentiated
Possible syndrome
16 Peptide receptors
Conclusions The molecular pathogenesis of neuroendocrine tumours of the gastrointestinal tract, to date remains largely unknown. Molecular mechanisms involved in the tumour genesis of neuroendocrine compared to non-endocrine tumours such as colo-rectal and pancreatic cancer are different. However, the significant recent advances, involving localization methods, natural history and new treatments, have prompted the attention of many studies on the peculiar characteristics of these tumours that could improve the over-all knowledge on dNETs.
Acknowledgements This study was supported by an inter-university cofinanced program 9906218982-005 (1999) from the Italian Ministry for University and Scientific and Technological Research (MURST), by a grant for finalised research from the Italian Ministry for Health 2000 and Fondazione Italiana per le Malattie Digestive (FIMAD). References 1. Solcia E, Kloppel G, Sobin LH. Histological Typing of Endocrine Tumours. Berlin/Heidelberg: Springer-Verlag 2000. 2. Wang D-G, Johnston CF, Anderson N et al. Overexpression of the tumor suppressor gene p53 is not implicated in neuroendocrine tumour carcinogenesis. J Pathol 1995; 175: 397-401. 3. Bartz C, Ziske C, Wiedenmann B et al. p53 tumour suppressor gene expression in pancreatic neuroendocrine tumour cells. Gut 1996; 38: 403-9. 4. Beghelli S, Pelosi G, Zamboni G et al. Pancreatic endocrine tumours: Evidence for a tumour suppressor pathogenesis and for a tumour suppressor gene on chromosome 17p. J Pathol 1998; 186: 41-50. 5. Ebert MP, Hoffmann J, Scheider-Stock R et al. Analysis of K-ras gene mutations in rare pancreatic and ampullary tumours. Eur J Gastroenterol Hepatol 1998; 10: 1025-9.
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
Increasing evidence exists that growth and proliferations, as well as the functional properties of various tumour cells, are modulated by gastrointestinal peptides. The low availability of stable cultured neuroendocrine tumour cell lines still represent a limiting factor in the study of the specific function of these receptors expressed on neuroendocrine tumours. However, for vasoactive intestinal peptide receptors (VIP-1 and VIP-2) and most of all for somatostatin receptors (ssts), numerous studies attempting to elucidate basic mechanisms of receptor functions and clinical applications of specific peptide agonists have been produced during the last few years. VIP receptors have been demonstrated, mostly by [123I]-VIP receptor scintigraphy, to be overexpressed on tumour cells of many intestinal adenocarcinomas as colorectal, pancreatic and gastric cancers, and in most of primary and metastatic digestive neuroendocrine tumours [20]. Although some studies suggest that VIP can regulate the growth and function of tumour cells [21], to date, no specific synthetic agonists for VIP-1 and/or VIP-2 receptor subtypes are available for clinical use. Therefore, radio-iodinated VIP for scintigraphic tumour visualization is the only clinical application available until now. On the contrary, during these years, studies on somatostatin and its receptor family are literally exploding. Many are the reasons accounting for the great interest that somatostatin has stimulated among basic and clinical scientists. Firstly: somatostatin mostly acts as an inhibitory factor that regulates a large number of physiological functions, such as inhibition of endocrine and exocrine secretions, modulation of neurotransmission, motor and cognitive function, inhibition of intestinal motility and intestinal smooth muscle cells contractility [22, 23], absorption of nutrients and ions, vascular contractility and cell proliferations [24, 25]. Secondly: during the last few years five different somatostatin receptor subtypes have been cloned and these are usually expressed in a cell tissue specific manner. Thirdly: synthetic somatostatin analogues, with relative subtype receptor specificity are available either for diagnostic (scintigraphy) and/or for therapeutic purposes in many diseases. Neuroendocrine tumours have been demonstrated to possess a high number of somatostatin receptors i.e., 100 times more than normal tissues. This property makes them a target of somatostatin agonists in the attempt to inhibit both secretion and growth by these tumour cells. In human digestive neuroendocrine tumour cells somatostatin and octreotide inhibit neurotransmitters and hormone secretions through inhibition of cellular calcium channels via sst2 and sst5 subtypes activation [26]. The antiproliferative action of somatostatin and its analogues (octreotide, lanreotide) induce Gi cell cycle arrest, activating distinct signal transduction pathways depending on receptor subtype. Sst2 antiproliferative effects are mediated by activation of tyrosine phosphatase SHP-1 and dephosphorylation of activated growth factor receptors (27); sstj mediates cell
growth arrest by stimulation of tyrosine phosphatase SHP-2, activation of MAP kinese pathway and induction of the cyclin-dependent kinase inhibitor p21 [28]; sst5 acts by a mechanism involving a dephosphorylation cascade, inhibition of cGMP-dependent protein kinase G and MAP kinase [29]. The antiproliferative effect of somatostatin can also result from apoptosis induced by sst3 subtype [30] or by indirect effects of the peptide resulting from the inhibition of secretion of growth promoting hormones and growth factors which specifically regulate tumour growth. These effects, although very impressive, have been mostly evaluated in experimental 'in vitro' conditions in somatostatin receptor subtypes transfected cells. However, more recent data on living patients with dNETs analysed prospectively indicate that the most beneficial antiproliferative effect of somatostatin analogues treatment account for 36% to 70% of tumour growth stabilization lasting from months to years [31, 32].
17 21. Pincus DW, DiCicco-Bloom EM, Blank IB et al. Vasoactive intestinal polypeptide regulates mitosis, differentiation and survival of cultured sympathetic neuroblasts. Nature 1990; 343: 56418. 22. Gu ZF, Corleto VD, Mantey SA et al. Somatostatin receptor subtype 3 mediates the inhibitory action of somatostatin on gastric smooth muscle cells. Am J Physiol 1995; 268: G739-G748. 23. Corleto VD, Severi C, Coy DH, et al. Colonic smooth muscle cells possess a different subtype of somatostatin receptor than gastric smooth muscle cells. Am J Physiol 1997; 272: G689-G697. 24. Patel YC. Molecular pharmacology of somatostatin receptor subtypes. J Endocrinol Invest 1997; 20: 348-67. 25. Benali N, Ferjoux G, Puente E et al. Somatostatin receptors. Digestion 2000; 62 (Suppl 1): 27-32. 26. Glassmeier G, Hopfner M, Riecken EO et al. Inhibition of L-type calcium channels by octreotide in isolated human neuroendocrine tumor cells of the gut. Biochem Biophys Res Commun 1998; 250: 511-5. 27. Bousquet C, Delesque N, Lopez F, et al. sst2 somatostatin receptor mediates negative regulation of insulin receptor signaling through the tyrosin phosphatases Shp-1. J Biol Chem 1998; 273: 7099-106. 28. FlorioT, Yao H, Carey KD. Somatostatin activation of mitogenactivated protein kinase via somatostatin receptorl (sstO. Mol Endocrinol 1999; 13: 24-37. 29. Cordelier P, Esteve JP, Bousquet C et al. Characterization of the antiproliferative signal mediated by the somatostatin receptor subtype 5. Proc Natl Acad Sci USA 1997; 94: 9343-8. 30. Sharma K, Srikant CB. Induction of wild-type p53, Bax, and acidic endonuclease durino somatostatin-signaled apoptosis in MCF-7 human breast cancer cells. Int J Cancer 1998; 76: 259-66. 31. Corleto VD, Angeletti S, Schillaci O, et al. Long-term octreotide treatment of metastatic carcinoid tumour. Ann Oncol 2000; 11: 491-3. 32. Arnold R, Simon B, Wied M. Treatment of neuroendocrine GEP tumours with somatostatin analogues. Digestion 2000; 62 (Suppl 1): 84-91.
Correspondence to: Gianfranco Delle Fave, MD Gastroenterology Department 11° Clinica Medica, Policlinico Umberto 1° 00161 Viale del Policlinico 155 Rome Italy E-mail: [email protected]
Downloaded from http://annonc.oxfordjournals.org/ at RMIT Central Library on July 1, 2015
6. Weber HC, Jensen RT. Pancreatic endocrine tumors and carcinoid tumors: recent insights from genetic and molecular biologic studies. In Dervenis CG (eds): Advances in Pancreatic Disease. Molecular Biology, Diagnosis and Treatment Stuttgart/New York: GeorgThiemeVerlag 1996; 55-75. 7. Canavese G, Azzoni C, Pizzi S et al. p27: A potential main inhibitor of cell proliferation in digestive endocrine tumors but not a marker of benign behavior. Hum Pathol 2000 (in press). 8. Bartsch D, Hahn SA, Danichevski KD et al. Mutation of the DPC4/Smad4 gene in neuroendocrine pancreatic tumors. Oncogenesl999; 18:2367-71. 9. Moore PS, Orlandini S, Zamboni G et al. Pancreatic tumours: Molecular pathways implicated in ductal cancer are involved in ampullary but not in esocrine nonductular or endocrine tumourigenesis. Cancer Res 2000 (in press). 10. Wang EH, Ebrahimi SA, Wu AYet al. Mutation of the MENIN gene in sporadic pancreatic endocrine tumors. Cancer Res 1998; 58: 4417-20. 11. Goebel SU, Heppner C, Burns AL et al. Genotype/phenotype correlation of multiple endocrine neoplasia type 1 gene mutations in sporadic gastrinomas. J Clin Endocrinol Metab 2000; 85: 11623. 12. Moore PS, Missiaglia E, Antonello D et al. The role of disease causing genes in sporadic pancreatic endocrine tumors. MEN1 and VHL. J Clin Endocrinol Metab 2000 (submitted). 13. Ruas M, Peters G. The pl61NK4a/CDKN2a tumor suppressor and its relatives. Biochim Biophys Acta 1998; 1378: F115-F117. 14. Muscarella P, Melvin WS, Fisher WE et al. Genetic alterations in gastrinomas and non-functioning pancreatic neuroendocrine tumors: An analysis of pl6/MTSl tumor suppressor gene inactivation. Cancer Res 1998; 58: 237^»0. 15. Serrano J, Goebel SU, Peghini PL et al. Alteration in the pl6 INK4a/CDK.N2A tumor suppressor gene in gastrinomas. J Clin Endocrinol Metab 2000; 85: 4146-56. 16. Wulbrand U, Wied M, Zofel P et al. Growth factor receptor expression in human gastroenteropancreatic neuroendocrine tumours. Eur J Clin Invest 1998; 28: 1038-^9. 17. Oberg K. Expression of growth factors and their receptors in neuroendocrine gut and pancreatic tumors, and prognostic factors for survival. Ann NYAcad Sd 1994; 733: 46-55. 18. Terris B, Scoazec JY, Rubbia L at al. Expression of vascular endothelial growth factor in digestive neuroendocrine tumours. Histopathology 1998; 32: 133-8. 19. Konno H, Arai T, TanakaTet al. Antitumor effect of neutralizing antibody to vascular endothelial growth factor on liver metastasis of endocrine neoplasm. Jpn J Cancer Res 1998; 89: 939-9. 20. Virgolmi I. Receptor nuclear medicine: Vasoactive peptide and somatostatin receptor scintigraphy for diagnosis and treatment of tumor patients. Eur J Clin Invest 1997; 27: 793-800.