Eur Radiol (2014) 24:2945–2952 DOI 10.1007/s00330-014-3317-4
COMPUTED TOMOGRAPHY
Pancreatic neuroendocrine tumours: correlation between MSCT features and pathological classification Yanji Luo & Zhi Dong & Jie Chen & Tao Chan & Yuan Lin & Minhu Chen & Zi-Ping Li & Shi-Ting Feng
Received: 28 January 2014 / Revised: 10 June 2014 / Accepted: 4 July 2014 / Published online: 22 July 2014 # European Society of Radiology 2014
M. Chen e-mail:
[email protected]
diagnosis of P-NENs, were included. Various MSCT features of the primary tumour, lymph node, and distant metastasis were analysed. The relationship between MSCT features and pathologic classification of P-NENs was analysed with univariate and multivariate models. Results Contrast-enhanced images showed significant differences among the three grades of tumours in the absolute enhancement (P = 0.013) and relative enhancement (P = 0.025) at the arterial phase. Univariate analysis revealed statistically significant differences among the tumours of different grades (based on World Health Organization [WHO] 2010 classification) in tumour size (P=0.001), tumour contour (P<0.001), cystic necrosis (P=0.001), tumour boundary (P=0.003), dilatation of the main pancreatic duct (P=0.001), peripancreatic tissue or vascular invasion (P<0.001), lymphadenopathy (P=0.011), and distant metastasis (P=0.012). Multivariate analysis suggested that only peripancreatic tissue or vascular invasion (HR 3.934, 95 % CI, 0.426–7.442, P= 0.028) was significantly associated with WHO 2010 pathological classification. Conclusions MSCT is helpful in evaluating the pathological classification of P-NENs. Key Points • P-NENs are potentially malignant, and classification of P-NENs carries important prognostic value. • MSCT plays an important role in the diagnosis and staging of P-NENs. • Correlations between classification of P-NENs and imaging findings have not been systematically evaluated. • MSCT could predict P-NENs classification and may be a useful prognostication tool.
Y. Lin Department of Pathology, The First Affiliated Hospital, Sun Yat-Sen University, 58th, The Second Zhongshan Road, Guangzhou, Guangdong, China 510080 e-mail:
[email protected]
Keywords Pancreatic neuroendocrine neoplasms . Multi-slice computed tomography (MSCT) . Pathological classification . Tumour . Prognosis
Abstract Objectives We aimed to evaluate the multi-slice computed tomography (MSCT) features of pancreatic neuroendocrine neoplasms (P-NENs) and analyse the correlation between the MSCT features and pathological classification of P-NENs. Methods Forty-one patients, preoperatively investigated by MSCT and subsequently operated on with a histological
Yanji Luo and Zhi Dong contributed equally to this work. Y. Luo : Z. Dong : Z.
2946
Eur Radiol (2014) 24:2945–2952
Introduction
MSCT protocol
P-NENs are considered rare, with an annual incidence of 1 in 100,000 individuals and accounting for approximately 1–2 % of all pancreatic tumours [1, 2]. The first pancreatic endocrine tumour was described in 1902 by Nicholls [3], who discovered the tumour incidentally at autopsy. In 1927, clinical hypersecretory syndrome caused by an insulinoma was first reported [4]. Since then, P-NENs have been categorized as functional or non-functional tumours depending on the presence or absence of clinical hormonal hypersecretion. The World Health Organization (WHO) 2010 classification [6] categorized P-NENs into neuroendocrine tumours grade 1 (G1) and grade 2 (G2), and neuroendocrine carcinomas grade 3 (G3), with large and small cell types based on the mitotic count and Ki-67 index. This classification also emphasized the malignant potential of all of these lesions [5, 6]. In addition, some clinical or pathological features of poorly differentiated P-NENs have been established, including large tumour size, irregular surface, high mitotic index, peritumoural vascular invasion, and distant metastasis [7]. To the best of our knowledge, there has been no prior study that attempted to systematically correlate pathological classification of P-NENs based on imaging findings. MSCT, given its panoramic capabilities and high spatial resolution, would appear to be a promising technique for predicting the classification of P-NENs. The aim of our study, therefore, was to identify MSCT findings of PNENs and to analyse the value of MSCT in the pathological classification of P-NENs.
All CT investigations were carried out on an MSCT scanner (Aquilion 64, Toshiba Medical Systems, Tokyo, Japan). Following the non-contrast imaging, iodinated contrast medium (Ultravist 300, Bayer Schering, Berlin, Germany) at a concentration of 300 mg iodine/ml was administered at a flow rate of 3–4 ml/s via a needle cannula with an automatic injector, which was placed in an antecubital vein, followed by a 40ml bolus of saline solution. Dual-phasic contrast-enhanced images were obtained during the arterial phase (33–35 s after initiation of the injection) and portal venous phase (60–65 s after initiation of the injection). The parameters for both noncontrast and contrast-enhanced CT examination were: tube voltage, 120 kV; tube current, 200–250 mAs, depending on the patient’s size; beam collimation, 64×0.5 mm; slice thickness, 0.5 mm; and rotation time, 0.4 s. MSCT features Qualitative analysis: All tumours were assessed for the following features: (1) tumour location: head, neck, body, or tail of the pancreas; (2) tumour shape: regular/round or irregular/ lobulated; (3) cystic-necrosis: a lesion was considered to show cystic necrosis if it contained areas of near-water attenuation or was entirely cystic-appearing, without enhancing tissue (Figs. 1 and 2); (4) tumour outline: well-defined or illdefined; (5) dilatation of the main pancreatic duct: positive if the diameter of the main pancreatic duct exceeded 3 mm, negative otherwise (Fig. 3); (6) peripancreatic tissue or vascular invasion: qualitative assessment based on ill-defined
Methods Patients and pathological classification The study was approved by the Institutional Review Board of Sun Yat-Sen University. From November 2009 to August 2013, all patients in our institution with pathologically confirmed sporadic P-NENs who had undergone MSCT imaging in the immediate preoperative period (within 14 days before surgery) were retrospectively evaluated. None of the patients had received any treatment at the time of imaging. Patients who had multiple endocrine neoplasia type 1 (MEN1 syndrome) or von Hippel-Lindau disease diagnosed in family history and confirmed genetic mutation were excluded from the study, as these are considered genetic rather than sporadic disease. All operative specimens were reviewed and classified by one pathologist. Histopathological classification of P-NENs was made according to WHO 2010 guidelines as: G1, mitotic count <2 per 10 high-power fields (HPF) and/or Ki-67≤2 %; G2, mitotic count 2–20 per 10 HPF and/ or Ki-67 3–20 %; and G3, mitotic count >20 per 10 HPF and/or Ki-67>20 %.
Fig. 1 Cystic-necrosis A 39-year-old woman with abdominal pain for more than one year was diagnosed with P-NENs (grade 2): contrastenhanced axial CT image showing a round pancreatic mass (white arrow) with clear edge and heterogeneous enhancement at the arterial phase, in which there was cystic-necrosis component
Eur Radiol (2014) 24:2945–2952
Fig. 2 A 68-year-old woman with epigastric pain for one week was diagnosed with P-NENs (grade 3): contrast-enhanced axial CT image demonstrating a well-defined cystic mass centred in the head of the pancreas, with peripheral enhancement (white arrow)
2947
Fig. 4 Invasion of peripancreatic tissue. A 58-year-old man with jaundice was diagnosed with P-NENs (grade 3): axial contrast-enhanced MSCT showing a heterogeneous soft-tissue mass of the pancreatic head/neck. Unclear boundary between the tumour and adjacent bowel can be seen (white arrows), indicating invasion of the adjacent tissues
boundaries between tumours and peripancreatic tissues/ vessels (Figs. 4 and 5); (7) lymphadenopathy: defined as the presence of lymph node(s) showing short axis diameter of more than 1 cm; and (9) distant metastasis (Fig. 6). Quantitative analysis: (1) primary tumour size: the longest axial diameter in either the axial, coronal, or sagittal plane, depending on the spatial orientation of the tumour (Fig. 7); (2) measurement of CT number (HU): the regions of interest (ROI) on the tumour (TROI) and normal pancreas (PROI) were drawn in the same axial image by one radiologist. The tumour ROI was drawn over tumours excluding the visible vessels, calcification, and the necrotic regions, and the size of the ROI varied from 5 to 15 mm in diameter. The CT values of TROI were measured at the pre-contrast phase, arterial phase, and portal venous phase, and were recorded as Tp, Ta, and Tv, respectively. The CT values of PROI were measured at the arterial phase and portal venous phase, and were recorded as
Pa and Pv, respectively. (3) The absolute enhancement of tumours at the arterial and portal venous phase (Tap and Tav, respectively), and relative enhancement between tumour and normal pancreas at the arterial and portal venous phase (TPa and TPv, respectively) were calculated using the following formulae:
Fig. 3 Dilatation of the main pancreatic duct. A 39-year-old man with abdominal pain of unknown reason was diagnosed with P-NENs (grade 2): contrast-enhanced axial CT image showing an irregular-shaped mass
tumour, which was detected in the pancreatic head region (a, white arrow), with proximal dilatation of the main pancreatic duct (b, white arrows) and associated atrophy of the pancreatic parenchyma
Tap ¼ Ta − Tp
ð1Þ
Tav ¼ Ta − Tv
ð2Þ
TPa ¼ Ta − Pa
ð3Þ
TPv ¼ Tv−Pv
ð4Þ
2948
Eur Radiol (2014) 24:2945–2952
Fig. 5 Peripancreatic vascular invasion. A 36-year-old woman with recurrent abdominal pain was diagnosed with P-NENs (grade 3). The axial contrast-enhanced imaging obtained during the arterial phase shows a large mass of the pancreatic head/neck surrounding the coeliac trunk and its branches (white arrows).
Fig. 7 Tumour size. A 35-year-old woman with repeated unconsciousness was diagnosed with insulinoma (grade 1): a, contrast-enhanced axial CT image of arterial phase showing a round high-density nodule in the pancreatic head with clear edge (white arrow); b, haematoxylin-eosin staining (×200) showing the small- to medium-sized tumour cells, either round or oval-shaped, without obvious mitotic figures, and normal pancreatic tissue in the lower right region
Imaging analysis MSCT images were retrospectively reviewed by two radiologists with 10 and 8 years of experience in abdominal imaging, who were blinded to the clinical and pathological data of all of the patients. Discrepancies between the readers were resolved by consensus after joint re-evaluation of the images. The examinations were reviewed in random order, with a time interval of at least one month and a mean interval of 21 days. Statistical analysis
Fig. 6 Liver metastasis. A 41-year-old man with abdominal pain for more than one year was diagnosed with P-NENs (grade 3): a, contrastenhanced CT showing low-attenuation mass in the pancreas (black arrow) with peripheral enhancing liver metastasis (white arrows); b, haematoxylin-eosin (×100) staining of the same P-NENs showing high cellularity and mitotic figures
Clinical features (sex, age, clinically functional or nonfunctional symptoms) and conventional MSCT features (tumour location, primary tumour size, tumour contour, cystic necrosis, tumour boundary, dilatation of main pancreatic duct, peripancreatic tissue or vascular invasion, lymphadenopathy, and distant metastasis) were analysed. In univariate analysis,
Eur Radiol (2014) 24:2945–2952
the continuous variables were reported as mean±standard deviation, and analysed using the Mann-Whitney U test, while the categorical variables were presented as numbers and percentages, and compared via Kruskal–Wallis test or Wilcoxon rank-sum test. The roles of these CT features were evaluated by multivariate ordinal regression to predict pathological classification. Comparison of enhanced MSCT values was performed with analysis of variance (ANOVA) or Kruskal–Wallis test according to the homogeneity of variance test. All of the statistical analyses were performed in SPSS (Version 18.0). All tests were two-sided, and a P value less than 0.05 was considered statistically significant.
Results Study population Between November 2009 and August 2013, there were 44 patients with pathologically proven P-NENs. Among the patients, one was diagnosed as having MEN1 syndrome, with multiple pancreatic tumours, and was thus excluded from our study. Another two patients were excluded because they underwent a surgical procedure more than 14 days (18 days and 23 days, respectively) after the MSCT examinations. Ultimately, therefore, a total of 41 patients, including 21 men (51 %) and 20 women (49 %), were included in the study. At diagnosis, the mean age was 48±14 years. Abdominal pain was observed in 11 patients (27 %), hypoglycaemia in nine (22 %), weight loss in seven (17 %), and obstructive jaundice in four (10 %), while an incidental diagnosis was reported in four cases (10 %) during health check. Other symptoms included paroxysmal confusion in three patients (7 %), palpitations in two (5 %), and watery stools in one (2 %). Based on the WHO 2010 classification, there were 21 cases of G1, eight cases of G2, and 12 cases of G3 tumours. There were 15 cases (37 %) of functional P-NENs, including 13 insulinomas, one gastrinoma, and one somatostatinoma. The general information for all of the patients is summarized in Table 1. Relationship between conventional MSCT findings of P-NENs and pathologic classification As shown in Table 2, all lesions were correctly identified and located with MSCT. The tumours originated from pancreatic head (n=18), neck (n=4), body (n=8), and tail (n=11). The mean size of the primary tumours was 3.3±2.3 cm (range, 0.6– 11 cm). Presence of cystic necrosis was suggested on MSCT in 24 (59 %) of the 41 patients. In three cases, enhancement was visible only in the peripheral portions of the lesion owing to their cystic nature. Dilation of the main pancreatic duct was
2949
positive in 13 (32 %) and negative in 28 cases (68 %). Peripancreatic tissue or vascular invasion was suggested in 14 cases (34 %) and excluded in 27 cases (66 %). Fifteen (37 %) of the 41 patients in the study had distant metastases: 13 with metastases only to the liver, one with bone metastases, and one with metastases in both the liver and the right adrenal gland. Univariate analysis revealed statistically significant difference among G1, G2, and G3 in tumour size and shape, cystic necrosis, tumour outline, dilatation of main pancreatic duct, peripancreatic tissue or vascular invasion, lymphadenopathy, and distant metastasis. There were no statistically significant differences in tumour location among the three groups (Table 2). An ordinal logistic regression model was used for multivariate analysis. The results suggested that only peripancreatic tissue or vascular invasion (HR 3.934, 95 % CI, 0.426–7.442, P=0.028) was significantly associated with pathological classification. The other factors were insignificant in multivariate analysis.
Relationship between enhancement and pathologic classification of P-NENs On unenhanced imaging, 31 (76 %) lesions were predominantly isodense to the pancreas, one (2 %) with intratumoural calcification was hyperdense, and nine (22 %) were hypodense (due to cystic or necrotic appearance). On contrast-enhanced scans, 29 (71 %) lesions were hypervascular compared to the normal pancreas, 10 (24 %) were hypovascular, and two (5 %) were isovascular (identified by the mass effect on MSCT scans). Both of the two isovascular cases were functional insulinomas. In the subgroup of 29 hypervascular lesions, maximum enhancement relative to the gland parenchyma was reached at the arterial phase in 20 patients (69 %) and the portal venous phase in nine patients (31 %). Both of the two isovascular cases were isodense on unenhanced imaging; they were identified based on mass effect. Both of the isovascular lesions were functional tumours. In this study, the average size of functional tumours was 18.5 mm and of non-functional P-NENs was 59.4 mm. The difference in size between these two groups was statistically significant (P=0.001). Other differences in imaging findings between the functional and non-functional tumours had no statistical significance. We found that both the absolute and relative enhancement of tumours decreased with higher pathological grading (Fig. 8). Since the parameters of Tap were heteroscedastic, the Kruskal–Wallis test was used to analyse the Tap parameters. As the parameters of Tav,TPa, and TPv were homoscedastic, these three parameters were analysed with ANOVA. The results showed significant differences among the three groups in Tap (P=0.013) and TPa (P=0.025) at the arterial phase. There was no significant difference among the three
2950
Eur Radiol (2014) 24:2945–2952
Table 1 General information of the patients with P-NENs 2
Z/x * P*
Pathological classification
Sex Male Female Age Functional Non-functional
G1
G2
G3
Total
8 13 48±15 11 10
5 8 21 (51 %) 1.671 0.095 3 4 20 (49 %) 48±11 50±15 48±14 0.071 0.932 2 2 15 (37 %) 1.683 0.092 6 10 26 (63 %)
Note: P-NENs=pancreatic neuroendocrine neoplasms; G1=grade 1; G2=grade 2; G3=grade 3; Functional=functional P-NENs with presence of clinical hormonal hypersecretion; Non-functional=non-functional PNENs with absence of clinical hormonal hypersecretion *Mann–Whitney U test was used to analyse the age of the patients. Sex and tumour function were analysed with the Wilcoxon rank-sum test
groups in Tav (P=0.198) or TPv (P=0.223) at the portal venous phase.
Discussion Although rare, P-NENs have been increasingly reported in the last few decades [8, 9].The WHO 2010 classifications are currently well-accepted by clinicians as prognostic factors of Table 2 Relationship between CT features and pathological classification of P-NENs
P-NENs. Among imaging modalities, MSCT plays an important role in the diagnosis and staging of P-NENs [10, 11]. Foti et al. reported a slight advantage of MSCT over MRI in evaluating the staging of P-NENs, particularly with regard to local invasion [12]. Large tumour size (<2 cm), local invasion, and lymph node enlargement tended to be associated with malignant P-NENs [13, 14]. All of the above studies were designed to differentiate between benign and malignant PNENs rather than investigating the correlation between imaging features and grading of pathologic classification. In this study, we systematically evaluated the non-contrast and contrast-enhanced MSCT features of P-NENs and attempted to identify any correlation with pathological classification. Our results indicated that imaging features are predictive of P-NEN classifications. They also showed that tumour enhancement at the arterial phase was correlated with tumour classification, and thus may be of prognostic value. Univariate analysis suggested that large tumour size, cystic necrosis, irregular tumour contour, unclear boundary, dilatation of the main pancreatic duct, peripancreatic tissue or vascular invasion, lymphadenopathy, and distant metastasis were more likely to occur in P-NENs of a higher grade. This is likely due to the fact that the pathological grading criteria of PNENs were established based on the mitotic count and Ki-67 index, both of which reflect the proliferation and invasiveness of tumour cells [6]. In general, we found that the poorer the differentiation and higher the grade of the tumour, the more
CT features
Location
Size
Tumour shape Cystic-necrosis Tumour outline Note: P-NENs=pancreatic neuroendocrine neoplasms; G1=grade 1; G2=grade 2; G3=grade 3 *Kruskal–Wallis test was used to analyse the tumour location and tumour size (statistical value is x2 ), and Wilcoxon rank-sum test was used to analyse other variables (statistical value is Z)
Dilatation of main pancreatic duct Peripancreatic tissue or vascular invasion Lymphadenopathy Distant metastasis
Pathological classification
Head Neck Body Tail ≤ 2 cm 2–4 cm >4 cm Round Lobular Present Absent Well-defined Ill-defined Present Absent Present Absent Present Absent Present Absent
G1
G2
G3
Total
11 4 3 3 13 8 0 18 3 7 14 17 4 2 19 1 20 6 15 5 16
1 0 1 6 1 4 3 5 3 6 2 6 2 3 5 2 6 4 4 1 7
6 0 4 2 1 6 5 2 10 11 1 3 9 8 4 11 1 9 3 9 3
18 4 8 11 15 18 8 25 16 24 17 26 15 13 28 14 27 19 22 15 26
χ2/Z*
P*
5.120
0.163
14.885
0.001
3.761
<0.001
3.377
0.001
2.950
0.003
3.360
0.001
4.773
<0.001
2.534
0.011
2.520
0.012
Eur Radiol (2014) 24:2945–2952
2951
Fig. 8 Box plots showing that difference in the MSCT enhancement of P-NENs of different pathologic grade: a, the absolute enhancement of P-NENs at the arterial (Tap, P=0.013) and portal venous phase (Tav, P= 0.198); b, the relative enhancement between tumour and normal pancreas at the arterial (TPa, P=0.025) and portal venous phase (TPv, P=0.223)
mitotic the phase and greater the aggressiveness. Our findings also corresponded with those of the study by Buetow et al., in which it was observed that large tumour size and presence of areas of cystic necrosis, as well as vascular invasion and distant metastases, tended to be seen in higher-grade tumours [13–16]. Bettini et al., however, reported that the main pancreatic ductal involvement did not correlate with the tumour grade of P-NENs, which was contrary to our results [13]. The discrepancy may be partly due to the fact that most of the tumours in our study were located in the pancreatic head, and therefore were more likely to involve the main pancreatic duct. The most significant novel finding of our study was that, with multivariate analysis and an ordinal logistic regression model, only peripancreatic tissue or vascular invasion was significantly associated with the pathological classification of P-NENs. The fact that this was more likely to be present in P-NENs of a higher grade was likely due to higher aggressiveness of the tumour cells. There was no significant association detected between the pathological classification and the other features in multivariate analysis. Such lack of association may be due to the interaction of these factors with peripancreatic tissue or vascular invasion, which may suggest that peripancreatic tissue or vascular invasion carries higher predictive value than other MSCT features in evaluating the pathologic classification of P-NENs. Similarly, Ekeblad et al. had proposed that vascular invasion was a more ominous prognostic factor than lymph node metastases in pancreatic endocrine tumours [17]. Most P-NENs are hyperattenuating on arterial-phase images, which is due to their rich capillary network. In our study, both absolute and relative enhancement of tumours decreased with higher tumour grade (Fig. 8). There were statistically significant differences among the three groups of P-NENs in both Tap and TPa, suggesting that the arterial blood supply of lower-grade or well-differentiated tumours was better than
that of higher-grade or poorly differentiated tumours. Gaspard et al. also found significant differences of blood flow and microvascular density (MVD) between benign tumours and tumours of uncertain behaviour or well-differentiated carcinomas [18, 19]. Blood flow was significantly higher in tumours with a Ki-67 proliferation index of 2 % or less. The CT perfusion parameters such as blood flow and mean transit time had good correlation with the classification of P-NENs [20]. Findings suggest that the greater the vascularization, the lower the grade of P-NENs. At the portal venous phase, Tav and TPv value also decreased as the tumour grade increased, suggesting that the venous reflux was less severe in lower-grade P-NENs. A possible explanation lies in the greater tendency of vessel invasion and microthrombus formation in poorly differentiated tumours, resulting in reduced artery supply and venous reflux compared to well-differentiated tumours [13, 21]. As there was no significant difference among the three tumour groups for Tav (P=0.198) and TPv (P=0.223) at the portal venous phase, it may be inferred that tumour enhancement during the arterial phase is more important than during the portal venous phase in predicting the classification of P-NENs. Gaspard et al. reported that perfusion CT measurements may be of prognostic value [18]. Our study may have been limited by the lack of functional data. While we analysed the enhancement characteristics of P-NENs by calculating the absolute and relative enhanced value of P-NENs on dualphase contrast-enhanced MSCT imaging, we were able to choose only two time points (arterial and portal venous phases), which may not represent the real-time dynamic enhancement pattern of the tumours. Dynamic contrastenhanced imaging could help to more comprehensively evaluate the enhancement pattern of the tumour. Another limitation was the lack of a bolus tracking technique in abdominal enhanced CT, which may affect the evaluation of vascularization of pancreatic lesions.
2952
In conclusion, our study showed that MSCT imaging is a feasible technique for predicting the pathological classification of P-NENs. The MSCT features of peripancreatic tissue or vascular invasion and lesser enhancement at the arterial phase were significantly correlated with higher grade of PNENs. MSCT may also prove to be useful in providing prognostic information with regard to patient outcome and in helping to monitor response to therapy. Acknowledgments The scientific guarantors of this publication are Dr. Shi-Ting Feng and Dr. Zi-Ping Li. The authors of this manuscript declare no relationships with any companies whose products or services may be related to the subject matter of the article. This study received funding from the following sources: National Natural Science Foundation of China (81000626), Zhujiang Scientific and Technological New Star Foundation (2012 J2200084), Natural Science Foundation of Guangdong Privince (S2013010016004) and Fundamental Research Funds for the Central Universities of China (10ykpy11). Lianxiong Yuan (expert in statistics, Sun Yat-Sen University) kindly provided statistical advice for this manuscript. Institutional Review Board approval was obtained, and written informed consent was waived by the Institutional Review Board. No study subjects or cohorts have been previously reported. Methodology: retrospective diagnostic or prognostic study, performed at one institution.
References 1. Oberg K, Eriksson B (2005) Endocrine tumours of the pancreas. Best Pract Res Clin Gastroenterol 19:753–781 2. Metz DC, Jensen RT (2008) Gastrointestinal neuroendocrine tumors: pancreatic endocrine tumors. Gastroenterology 135:1469–1492 3. Nicholls AG (1902) Simple Adenoma of the Pancreas arising from an Island of langerhans. J Med Res 8:385–395 4. Wilder R, Allan F, Power M, Robertson H (1927) Carcinoma of the islands of the pancreas: hyperinsulinism and hypoglycemia. J Am Med Assoc 89:348–355 5. Casadei R, Ricci C, Pezzilli R et al (2009) Value of both WHO and TNM classification systems for patients with pancreatic endocrine tumors: results of a single-center series. World J Surg 33:2458–2463 6. Rindi G, Arnold R, Bosman F et al (2010) Nomenclature and classification of neuroendocrine neoplasms of the digestive system. In: Bosman F, Carneiro F, Hruban R, Theise N (eds) WHO classification of tumors of the digestive system. IARC Press, Lyon, pp 13–14 7. Zerbi A, Falconi M, Rindi G et al (2010) Clinicopathological features of pancreatic endocrine tumors: a prospective multicenter study in Italy of 297 sporadic cases. Am J Gastroenterol 105:1421–1429
Eur Radiol (2014) 24:2945–2952 8. Yao JC, Hassan M, Phan A et al (2008) One hundred years after "carcinoid": epidemiology of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol 26: 3063–3072 9. Panzuto F, Boninsegna L, Fazio N et al (2011) Metastatic and locally advanced pancreatic endocrine carcinomas: analysis of factors associated with disease progression. J Clin Oncol 29:2372–2377 10. Rappeport ED, Hansen CP, Kjaer A, Knigge U (2006) Multidetector computed tomography and neuroendocrine pancreaticoduodenal tumors. Acta Radiol 47:248–256 11. Bettini R, Boninsegna L, Mantovani W et al (2008) Prognostic factors at diagnosis and value of WHO classification in a mono-institutional series of 180 non-functioning pancreatic endocrine tumours. Ann Oncol 19:903–908 12. Foti G, Boninsegna L, Falconi M, Mucelli RP (2013) Preoperative assessment of nonfunctioning pancreatic endocrine tumours: role of MDCT and MRI. Radiol Med 118:1082–1101 13. Bettini R, Partelli S, Boninsegna L et al (2011) Tumor size correlates with malignancy in nonfunctioning pancreatic endocrine tumor. Surgery 150:75–82 14. Gallotti A, Johnston RP, Bonaffini PA et al (2013) Incidental neuroendocrine tumors of the pancreas: MDCT findings and features of malignancy. AJR Am J Roentgenol 200:355–362 15. Buetow PC, Parrino TV, Buck JL et al (1995) Islet cell tumors of the pancreas: pathologic-imaging correlation among size, necrosis and cysts, calcification, malignant behavior, and functional status. AJR Am J Roentgenol 165:1175–1179 16. Lewis RB, Lattin GE Jr, Paal E (2010) Pancreatic endocrine tumors: radiologic-clinicopathologic correlation. Radiographics 30:1445– 1464 17. Ekeblad S, Skogseid B, Dunder K, Oberg K, Eriksson B (2008) Prognostic factors and survival in 324 patients with pancreatic endocrine tumor treated at a single institution. Clin Cancer Res 14:7798– 7803 18. Rodallec M, Vilgrain V, Couvelard A et al (2006) Endocrine pancreatic tumours and helical CT: contrast enhancement is correlated with microvascular density, histoprognostic factors and survival. Pancreatology 6:77–85 19. d’Assignies G, Couvelard A, Bahrami S et al (2009) Pancreatic endocrine tumors: tumor blood flow assessed with perfusion CT reflects angiogenesis and correlates with prognostic factors. Radiology 250:407–416 20. Couvelard A, O'Toole D, Turley H et al (2005) Microvascular density and hypoxia-inducible factor pathway in pancreatic endocrine tumours: negative correlation of microvascular density and VEGF expression with tumour progression. Br J Cancer 92: 94–101 21. Marion-Audibert AM, Barel C, Gouysse G et al (2005) Low microvessel density is an unfavorable histoprognostic factor in pancreatic endocrine tumors. Gastroenterology 125:1094–1104