Cytological Ki-67 in pancreatic endocrine tumours: an opportunity for pre-operative grading

Transcription

1 Cytological Ki-67 in pancreatic endocrine tumours: an opportunity for pre-operative grading Cecilia Piani, Giulia M Franchi, Chiara Cappelletti, Marina Scavini, Luca Albarello 1, Alessandro Zerbi 2, Paolo Giorgio Arcidiacono 3, Emanuele Bosi and Marco F Manzoni General Medicine, Diabetes and Endocrinology Unit, San Raffaele Scientific Institute, Università Vita-Salute San Raffaele, Via Olgettina 60, Milano, Italy 1 Pathology Unit, 2 Pancreatic Unit, Department of Surgery, and 3 Gastroenterology and Endoscopic Unit, San Raffaele Scientific Institute, Università Vita-Salute San Raffaele, Milano, Italy (Correspondence should be addressed to M F Manzoni; manzoni.marco@hsr.it) C Piani and G M Franchi contributed equally to this work Abstract The cytological Ki-67 expression measured on cytological samples collected by endoscopic ultrasonography-guided fine needle aspiration cytology (EUS-FNAC) may provide pre-operative indications for pancreatic endocrine tumours (PETs) management. The aim of our study was to assess reliability of Ki-67 expression measured on cytological samples obtained by EUS-FNAC in patients with PETs. Eighteen patients with PETs underwent EUS-FNAC before surgery. Ki-67 expression was measured on FNACs and on histological sections. Using a cut-off of 2%, percent agreement of Ki-67 expression on cytological and histological samples was 89% (k-statistic: 0.78, 95% confidence intervals (95% CI): 0.5, 1.0). Using cut-off values of 2 and 10%, percent agreement was 78% (k-statistic: 0.65, 95% CI: 0.3, 0.9). Ki-67 expression measured on cytological samples obtained by EUS-FNAC before surgery showed good agreement with that measured on histological samples. Introduction Neuroendocrine tumours (NETs) are a heterogeneous group of rare neoplasms that account for 0.5% of all malignancies, with an incidence of two new cases per people per year in the US. The increasing incidence of NETs observed in the last few decades may be accounted for by increased awareness, improved diagnostic tools and a revision in the definition. The main primary sites of NETs are the gastro entero pancreatic (GEP) tract (62 67%) and the lung (22 27%; Taal & Visser 2004). In patients with GEP NETs, the strongest predictor of 5-year survival is the staging (93% for local disease, 74% for regional disease and 19% for metastatic disease). Therefore, early diagnosis of GEP NETs is a crucial issue (Arnold 2005). GEP NETs include pancreatic endocrine tumours (PETs). PETs occur in adults of all age groups, with no sex differences (prevalence of w1 in individuals) and are extremely rare in children. PETs may be isolated or occur in association with other endocrine tumours in the multiple endocrine neoplasia type 1 (MEN-1) syndrome or, more rarely, in the von Hippel Lindau syndrome (Mansour & Chen 2004, Oberg & Eriksson 2005). PETs, like all NETs, are categorized on the basis of their clinical manifestations into functioning and non-functioning tumours. The functioning tumours are associated with a clinical syndrome caused by the inappropriate secretion of hormones, such as insulin, gastrin, vasoactive intestinal polypeptide, glucagon or other even less common hormones, such as adrenocorticotrophic hormone or growth hormone (Kloppel et al. 2004). PETs are usually macroscopically well demarcated, solitary, round tumours with a diameter between 1 and 4 cm and they occur in all pancreatic segments /08/ q 2008 Society for Endocrinology Printed in Great Britain Downloaded from Bioscientifica.com DOI: /ERC at 04/19/ :56:12PM Online version via

2 C Piani, G M Franchi et al.: Cytological Ki-67 in PETs According to the World Health Organization (WHO) classification, PETs can be categorized as well-differentiated endocrine tumours (with benign or uncertain biological behaviour), well-differentiated endocrine carcinomas and poorly differentiated endocrine carcinomas. The pathological criteria of malignancy are lymph node invasion, distant metastases and invasion of adjacent organs; tumour size O2 cm in diameter, angioinvasion and proliferative activity of more than 2% are the criteria that distinguish welldifferentiated endocrine tumours with uncertain biological behaviour from well-differentiated endocrine tumours with benign biological behaviour (De Lellis et al. 2004). Prior to surgery, organ infiltration and the presence of distant metastasis are the only criteria that clinicians have with which to assess the malignancy of PETs. However, these features are present in no more than 30 50% of malignant PETs at diagnosis, and, therefore, a reliable prediction of malignancy is usually not available prior to surgery (Pelosi et al. 1996). The Ki-67 antigen is present in all proliferating cells and it can be detected during all active phases of the cell cycle. Therefore, its expression is a reliable measure of the growth fraction of a given cell population (Scholzen & Gerdes 2000, Brown & Gatter 2002). Many studies have shown the prognostic value of Ki-67 measured on the surgically removed tumour: a low expression of Ki-67 indicates a slow tumour growth and a favourable course of the disease, whereas high Ki-67 expression predicts a poor prognosis owing to more rapid tumour growth (Pelosi et al. 1992, 1996, Gentil Perret et al. 1998, Rindi et al. 1998, Rosenau et al. 2002, Oberg 2003, Vilar et al. 2007). In fact, Ki- 67 is one of the criteria used to grade PETs according to the WHO classification (De Lellis et al. 2004). Although recent studies have shown that the expression of cytokeratin 19 (CK 19) predicts long-term survival among patients with PETs (Deshpande et al. 2004, La Rosa et al. 2007), this marker is not yet widely used in the clinical setting. The precise localization of PETs is of utmost importance because surgical resection is the curative treatment. However, preoperative localization can be difficult, as these tumours are frequently smaller than 2 cm in diameter. Endoscopic ultrasound is increasingly used in the preoperative localization and diagnosis of pancreatic tumours, including PETs, as it provides high-resolution images of structures within or just beyond the wall of the gastrointestinal tract, which allows the detection of pancreatic lesions down to cm (Anderson et al. 2000, Chang et al. 2006). Furthermore, fine needle aspiration cytology 176 (FNAC) of pancreatic masses can be performed during EUS, with a low rate of complications allowing the preoperative cytological analysis and characterization of the lesions (McLean & Fairclough 2005, Chang et al. 2006). The aim of this study was to assess reliability of the cytological Ki-67 measured on cytological samples obtained during EUS-FNAC in patients with PETs; we compared the measurements of the cytological Ki-67 expression obtained on EUS-FNAC and the histological Ki-67 expression obtained on histological sections after surgical removal of the tumour. Patients In this study, we included 24 patients referred to our institute from October 2003 to January 2007 for evaluation of PETs. All patients underwent CT and EUS-FNAC examination before surgery for the removal of the pancreatic mass. Ki-67 expression was measured by immunocytochemistry on FNACs and on the corresponding histological sections by immunohistochemistry. All patients signed an informed consent before the procedure. In 18 patients (nine men and nine women; mean age at diagnosis 53.9 years; range 24 82), we measured Ki-67 expression on both the cytological and histological samples. The pathologist measuring Ki-67 expression on histological sections was blinded to the results of the cytology. In six patients we could not measure Ki-67 expression on the cytological sample because the sample was either not collected for technical reasons or not enough material was left after performing routine staining. The diagnosis of PET was confirmed by the histological examination of the removed tissue. The functional status of tumours was assessed on the presence of a clinical syndrome and confirmed with the determination of plasma concentrations of specific hormones. Patients with a history of hypoglycaemia or symptoms suggestive of hypoglycaemia underwent a 72-h fasting test during which plasma glucose, insulin, pro-insulin and C-peptide were drawn every 6 h or whenever hypoglycaemic symptoms occurred. The test was stopped if plasma glucose was!2.8 mmol/l or if the patient had hypoglycaemic symptoms. The fasting test was consistent with the presence of an insulinoma/pro-insulinoma if, concomitant to the hypoglycaemia, we recorded a plasma insulin concentration O3 mu/ml or a plasma pro-insulin concentration O15 pmol/l by immunochemiluminometric assay, and a C-peptide O0.33 nmol/l.

3 PETs without clinical manifestation were classified as non-functioning tumours; they were often associated with increased levels of chromogranin A and neuronspecific enolase and they had the typical features of endocrine tumours on imaging. EUS-FNAC technique Patients were sedated with i.v. propofol (Diprivan; AstraZeneca, Basiglio, Milan, Italy). EUS-FNAC was performed using a linear Pentax EG 3630UX or EG 3830UT (Pentax Precision Instruments Corp., Orangeburg, NY, USA) echoendoscope. The echoendoscope was advanced to the descending duodenum beyond the major ampulla and slowly withdrawn from the duodenum to the stomach. To ensure that all regions of the pancreas were visualized, regional anatomy was verified by its relationship to the surrounding vessels and organs. The tumour s size, echotexture, location within the pancreas, involvement of peripancreatic vessels and the presence of enlarged lymph nodes were documented. FNAC was performed under direct EUS guidance using 22- and 25-gauge needles (Wilson- Cook Medical Inc., Winston-Salem, NC, USA). Immunocytochemical procedures Direct smears were prepared, and alcohol-fixed samples were stained by haematoxylin eosin (H and E) for morphological evaluation. Cytological features assessed in the aspirated samples included cellularity, cellular cohesiveness, uniformity of the tumour cell population, location of the nuclei, nuclear pleomorphism, nuclear membrane irregularity, chromatin pattern, presence of nucleoli, cytoplasmic features, necrosis and the presence of additional cell types in the smears. Immunocytochemical stains were performed with antiserum against synaptophysin to confirm neuroendocrine differentiation and against Ki-67 to measure the proliferation index of neoplastic cells. Antibodies against synaptophysin (Snp88; Biogenex, San Ramon, CA, USA) and Ki-67 (MIB1; DakoCytomation, Carpineteria, CA, USA) were used at a 1/1200 and 1/600 dilution respectively. One representative sample from each case was treated with microwave for 5 min in citrate buffer solution (ph 6.0) for antigen retrieval. After endogenous peroxidase inhibition with 3% hydrogen peroxide solution for 5 min, cytological samples were rinsed in Tris-buffered saline (TBS), and reacted with a primary antibody for 60 min at room temperature. After rinsing with TBS, amplification was performed with a super enhancer (post primary) for 20 min and with a polymer for 30 min (Novocastra revelation kit; Vision BioSystems Novocastra, Newcastle upon Tyne, UK). The antigen antibody reaction was visualized using 10 ml 3,3 0 -diaminobenzidine. The percentage of Ki-67 reactive tumour cells was calculated under light microscope and immunoreactivity was considered positive in the presence of nuclear staining (Fig. 1). Immunohistochemical procedures Three-micrometre thin consecutive paraffin sections of formalin-fixed, paraffin-embedded tissues were placed on slides coated with poly-l-lysine. Immunohistochemical stains were performed with antibodies against synaptophysin and Ki-67 to determine neuroendocrine differentiation and the proliferation index (see above). After deparaffinization and blocking of endogenous peroxidase, antigen retrieval treatment was performed. Samples were rinsed in TBS and reacted with a primary antibody for 60 min at room temperature. After rinsing with TBS, amplification was performed with a super enhancer (post primary) for 20 min and with a polymer for 30 min (Novocastra revelation kit, Vision BioSystems Novocastra). The antigen antibody reaction was visualized using 10 ml 3,3 0 -diaminobenzidine. The value of Ki-67 was obtained counting the fraction of positive cells and expressing this result as a percentage (Fig. 2). Statistical analysis Proportions are presented with 95% confidence intervals (95% CI) in parentheses. For the statistical analysis, Ki-67 expression!1% was considered as 1%. We categorized Ki-67 expression using as a cutoff 2% (categories: Ki-67%2% and Ki-67O2%; De Lellis et al. 2004), or using cut-off of 2 and 10% (categories: Ki-67%2%, 2%!Ki-67%10% and Ki-67O10%; Rindi et al. 1998, De Lellis et al. 2004, Figure 1 Ki-67 on a PET s cytological smear obtained with EUS-FNAC

4 C Piani, G M Franchi et al.: Cytological Ki-67 in PETs A B C Figure 2 Ki-67 on a PET s histological section obtained after surgical removal of the tumour: original magnification 200! (panel A) or 400! (Panel B). Normal pancreatic tissue is provided for comparison in panel C (original magnification 400!). Vilar et al. 2007). Agreement between Ki-67 expression measured by immunocytochemistry and immunohistochemistry was assessed using percent agreement and k-statistics (Landis & Koch 1977). The k-statistic was interpreted according to Altman (1991). Statistical significance was P! Results According to the WHO classification, six patients had a well-differentiated endocrine tumour (WDET) with benign biological behaviour, seven had a WDET with uncertain biological behaviour and five had a welldifferentiated endocrine carcinoma (WDEC). Eleven tumours were non-functioning PETs, three were insulinomas and four were pro-insulinomas. Of the 18 cases, 7 pancreatic masses were located in the head of the pancreas, 7 in the body and 4 in the tail. Nine tumours had a diameter!2 cm. None of the patients who underwent EUS-FNAC examination developed side effects due to this procedure. Results of the Ki-67 expression measured by immunocytochemistry and immunohistochemistry are reported in Table 1. When using a cut-off of 2%, Ki-67 expression measured by immunocytochemistry and immunohistochemistry was concordant in 16 out of 18 cases (eight cases in the %2% category and eight in the O2% category) and the remaining 2 cases were discordant (Ki-67 expression by immunocytochemistry 9 and 3%; Ki-67 expression by immunohistochemistry 1 and!1% respectively; see Table 2). The percent agreement between Ki-67 expression by immunocytochemistry and immunohistochemistry was 89% (95% CI: 65.3, 98.6). The k-statistic was 0.78 (95% CI: 0.5, 1.0; PZ0.0003), suggesting a good agreement (Altman 1991). When using cut-off of 2 and 10%, Ki-67 expression measured by immunocytochemistry and by immunohistochemistry was concordant in 14 out of 18 cases (eight cases in the %2% category, three cases in the 2%!Ki-67%10% category and three cases in the O10% category) and the remaining 4 cases were discordant (Ki-67 expression by immunocytochemistry 18, 5, 9 and 3%; Ki-67 expression by immunohistochemistry 10, 12, 1 and!1% respectively; see Table 3). The percent agreement between Ki-67 expression by immunocytochemistry and immunohistochemistry was 78% (95% CI: 52.4, 93.6). The k-statistic was 0.65 (95% CI: 0.3, 0.9; PZ0.0001), suggesting a good agreement (Altman 1991). Discussion Our study shows that among patients with PETs, it is feasible and safe to measure Ki-67 expression on cytological samples obtained by EUS-FNAC and that Ki-67 expression measured on cytological samples has a good agreement with the Ki-67 expression measured on histological sections.

5 Table 1 General characteristics of the patients and the pancreatic mass Case Sex Age WHO Type Site Diameter (mm) 1 F 55 WDEC Insulinoma Body 36 2 M 32 WDEC Pro-insulinoma Body 10 3 F 71 WDET a Non-functioning Head 13 4 M 64 WDET a Non-functioning Head 16 5 F 67 WDEC Non-functioning Tail 95 6 F 68 WDET a Insulinoma Tail 12 7 F 46 WDET a Pro-insulinoma Head 15 8 M 37 WDET a Non-functioning Head 18 9 M 65 WDEC Non-functioning Head F 31 WDET b Insulinoma Body F 75 WDET b Non-functioning Head M 72 WDEC Non-functioning Tail M 24 WDET b Pro-insulinoma Head M 36 WDET a Non-functioning Body F 69 WDET b Non-functioning Body F 82 WDET b Non-functioning Body M 45 WDET b Non-functioning Body M 31 WDET b Pro-insulinoma Tail 23 a WDET, well-differentiated endocrine tumours with benign biological behaviour. b WDET, well-differentiated endocrine tumours with uncertain biological behaviour. Although the literature reports discordant opinions on the value of tumour proliferation markers in predicting a patient s prognosis, many studies have shown that Ki-67 expression in histological sections obtained from PETs is an important predictor of their Table 2 Ki-67 expression measured by immunocytochemistry and immunohistochemistry in primary pancreatic endocrine tumours (PETs) measured on cellular material collected through endoscopic ultrasonography-guided fine needle aspiration cytology (EUS-FNAB) and on surgically removed tumour tissue Case Cytological Ki-67 (%) Histological Ki-67 (%) !1 4!1! !1!1 7!1!1 8!1! ! ! !1 biological behaviour (Pelosi et al. 1992, 1996, Rindi et al. 1998, Rigaud et al. 2001, Butturini et al. 2006, La Rosa et al. 2007, Vilar et al. 2007). The WHO classification of PETs includes Ki-67 expression in the list of parameters (e.g. distant metastases, organ infiltration, dimension, angio/neuroinvasion, number of mitosis) determining the patient s prognosis (De Lellis et al. 2004). Furthermore, some authors have demonstrated that Ki-67 index is associated with a patient s outcome, i.e. patients with pancreatic endocrine tumours with Ki-67 index O2% have a significant increased mortality compared with those with a Ki-67 index %2% (HRZ11.7 (95% CI 1.98, 72.3); Rigaud et al. 2001). The cytological samples obtained by EUS-FNAC may be used to study the Table 3 Ki-67 expression measured on histological and cytological samples, considering two (upper panel; De Lellis et al. 2004) or three (lower panel; Rindi et al. 1998, De Lellis et al. 2004, Vilar et al. 2007) Ki-67 categories Cytological Histological %2% O2% %2% 8 2 O2% 0 8 %2% 2! Ki-67%10% O10% %2% ! Ki-67%10% O10%

6 C Piani, G M Franchi et al.: Cytological Ki-67 in PETs expression of proliferation markers described in other types of tumours, such as E-cadherin (Rosenau et al. 2002) and p53 (Kawahira et al. 2000). A major limitation of this technique is sample collection, since in six patients the specimen collected by fine needle aspiration was inadequate for measuring Ki-67 expression. EUS-FNAC is a highly operatordependent technique and tumours in the lower head or in the uncinate process of the pancreas may be of difficult access (McLean & Fairclough 2005). Indeed, of the six cases with an inadequate sample three had a PET located in the head (nz2) or in the uncinate process (nz1) of the pancreas. However, we were successful in obtaining samples adequate for cytological analysis in other seven PETs located in the head of the pancreas. Another limitation of this technique is related to intratumoural heterogeneity and differential expression of Ki-67 in primary tumours and metastases (Chaudhry et al. 1992). When Ki-67 expression is measured on multiple histological sections of PETs, the highest Ki-67 expression is reported. Although Ki-67 expression measured on cytological samples may not completely reflect Ki-67 expression heterogeneity of PETs, our study showed a good agreement between categories of Ki-67 expression measured on histological sections or cytological samples. Infiltration of adjacent organs and distant metastasis were the only parameters defining the malignancy of PETs before tissues are collected for pathology during surgery. Since the EUS-FNAC is a relatively noninvasive procedure, Ki-67 expression measured on cytological samples could be easily obtained in all patients with PETs before surgery. The preoperative availability of Ki-67 expression, combined with number and size of lesions, site of the tumour, expression of somatostatin receptors, peripancreatic infiltration, presence of distant metastases or multiple tumours, patients performance status and clinical symptoms, may help the clinician choosing the best therapeutic approach. In the case of a single pancreatic lesion, the preoperative availability of Ki-67 expression may help in selecting the best surgical intervention between atypical resection (enucleation or middle pancreatectomy) and typical resection (pancreaticoduodenectomy or distal pancreatectomy). In patients with MEN-1 syndrome, who often have multiple pancreatic lesions arising at different times over the years (Oberg & Eriksson 2005), preoperative availability of Ki-67 expression may help with optimizing both extension and timing of surgery. However, longitudinal studies are needed to assess the usefulness of preoperative availability of Ki-67 in the management of patients with PETs. 180 In conclusion, this study demonstrates an overall good agreement between the expression of Ki-67 in primary PETs measured on cellular material collected through EUS-FNAC and on surgically removed tumour tissue. The cytological Ki-67 may improve the preoperative assessment of PETs, helping the clinician choosing the optimal therapeutical approach. Acknowledgements The authors wish to thank Prof. V Di Carlo, MD and C Doglioni, MD for their continuous support and valuable discussion of the data. The authors have no conflict of interest to disclose. References Altman DG 1991 Practical Statistics for Medical Research. London, England: Chapman and Hall. Anderson MA, Carpenter S, Thompson NW, Nostrant TT, Elta GH & Scheiman JM 2000 Endoscopic ultrasound is highly accurate and directs management in patients with neuroendocrine tumors of the pancreas. American Journal of Gastroenterology Arnold R 2005 Endocrine tumours of the gastrointestinal tract. Introduction: definition, historical aspects, classification, staging, prognosis and therapeutic options. Best Practice & Research. Clinical Gastroenterology Brown DC & Gatter KC 2002 Ki67 protein: the immaculate deception? Histopathology Butturini G, Bettini R, Missiaglia E, Mantovani W, Dalai I, Capelli P, Ferdeghini M, Pederzoli P, Scarpa A & Falconi M 2006 Predictive factors of efficacy of the somatostatin analogue octreotide as first line therapy for advanced pancreatic endocrine carcinoma. Endocrine-Related Cancer Chang F, Vu C, Chandra A, Meenan J & Herbert A 2006 Endoscopic ultrasound-guided fine needle aspiration cytology of pancreatic neuroendocrine tumours: cytomorphological and immunocytochemical evaluation. Cytopathology Chaudhry A, Oberg K & Wilander E 1992 A study of biological behavior based on the expression of a proliferating antigen in neuroendocrine tumors of the digestive system. Tumour Biology Deshpande V, Fernandez-del Castillo C, Muzikansky A, Deshpande A, Zukerberg L, Warshaw AL & Lauwers GY 2004 Cytokeratin 19 is a powerful predictor of survival in pancreatic endocrine tumors. American Journal of Surgical Pathology Gentil Perret A, Mosnier JF, Buono JP, Berthelot P, Chipponi J, Balique JG, Cuilleret J, Dechelotte P & Boucheron S 1998 The relationship between MIB-1 proliferation index and outcome in pancreatic neuroendocrine tumors. American Journal of Clinical Pathology

7 Kawahira H, Kobayashi S, Kaneko K, Asano T & Ochiai T 2000 p53 protein expression in intraductal papillary mucinous tumors (IPMT) of the pancreas as an indicator of tumor malignancy. Hepatogastroenterology Kloppel G, Perren A & Heitz PU 2004 The gastroenteropancreatic neuroendocrine cell system and its tumors: the WHO classification. Annals of the New York Academy of Sciences Landis JR & Koch GG 1977 The measurement of observer agreement for categorical data. Biometrics De Lellis R, Lloyd RV, Heitz PU & Eng C 2004 Pathology and Genetics of Tumours of Endocrine Organs (World Health Organization Classification of Tumours). Lyon: IARC Press. Mansour JC & Chen H 2004 Pancreatic endocrine tumors. Journal of Surgical Research McLean AM & Fairclough PD 2005 Endoscopic ultrasound in the localisation of pancreatic islet cell tumours. Best Practice and Research. Clinical Endocrinology and Metabolism Oberg K 2003 Diagnosis and treatment of carcinoid tumors. Expert Review of Anticancer Therapy Oberg K & Eriksson B 2005 Endocrine tumours of the pancreas. Best Practice and Research. Clinical Gastroenterology Pelosi G, Zamboni G, Doglioni C, Rodella S, Bresaola E, Iacono C, Serio G, Iannucci A & Scarpa A 1992 Immunodetection of proliferating cell nuclear antigen assesses the growth fraction and predicts malignancy in endocrine tumors of the pancreas. American Journal of Surgical Pathology Pelosi G, Bresaola E, Bogina G, Pasini F, Rodella S, Castelli P, Iacono C, Serio G & Zamboni G 1996 Endocrine tumors of the pancreas: Ki-67 immunoreactivity on paraffin sections is an independent predictor for malignancy: a comparative study with proliferating-cell nuclear antigen and progesterone receptor protein immunostaining, mitotic index, and other clinicopathologic variables. Human Pathology Rigaud G, Missiaglia E, Moore PS, Zamboni G, Falconi M, Talamini G, Pesci A, Baron A, Lissandrini D, Rindi G et al High resolution allelotype of nonfunctional pancreatic endocrine tumors: identification of two molecular subgroups with clinical implications. Cancer Research Rindi G, Capella C & Solcia E 1998 Cell biology, clinicopathological profile, and classification of gastro enteropancreatic endocrine tumors. Journal of Molecular Medicine La Rosa S, Rigoli E, Uccella S, Novario R & Capella C 2007 Prognostic and biological significance of cytokeratin 19 in pancreatic endocrine tumours. Histopathology Rosenau J, Bahr MJ, von Wasielewski R, Mengel M, Schmidt HH, Nashan B, Lang H, Klempnauer J, Manns MP & Boeker KH 2002 Ki67, E-cadherin, and p53 as prognostic indicators of long-term outcome after liver transplantation for metastatic neuroendocrine tumors. Transplantation Scholzen T & Gerdes J 2000 The Ki-67 protein: from the known and the unknown. Journal of Cellular Physiology Taal BG & Visser O 2004 Epidemiology of neuroendocrine tumours. Neuroendocrinology 80 (Suppl 1) 3 7. Vilar E, Salazar R, Perez-Garcia J, Cortes J, Oberg K & Tabernero J 2007 Chemotherapy and role of the proliferation marker Ki-67 in digestive neuroendocrine tumors. Endocrine-Related Cancer