World J Surg DOI 10.1007/s00268-017-4317-8

ORIGINAL SCIENTIFIC REPORT

Preoperative Imaging Overestimates the Tumor Size in Pancreatic Neuroendocrine Neoplasms Associated with Multiple Endocrine Neoplasia Type 1 V. Polenta1,5 • E. P. Slater1 • P. H. Kann2 • M. B. Albers1 • J. Manoharan1 • A. Ramaswamy3 • A. H. Mahnken4 D. K. Bartsch1

•

Ó Socie´te´ Internationale de Chirurgie 2017

Abstract Background Radiological tumor size of non-functioning pancreatic neuroendocrine neoplasms (Nf-pNENs) associated with multiple endocrine neoplasia type 1 (MEN1) is a crucial parameter to indicate surgery. The aim of this study was to compare radiological size (RS) and pathologic size (PS) of MEN1 associated with pNENs. Methods Prospectively collected data of MEN1 patients who underwent pancreatic resections for pNENs were retrospectively analyzed. RS was defined as the largest tumor diameter measured on endoscopic ultrasound (EUS), magnetic resonance imaging (MRI) or computed tomography (CT). PS was defined as the largest tumor diameter on pathological analysis. Student’s t test and linear regression analysis were used to compare the median RS and PS. p \ 0.05 was considered significant. Results Forty-four patients with a median age of 37 (range 10–68) years underwent primary pancreatic resections for pNENs. Overall, the median RS (20 mm, range 3–100 mm) was significantly larger than the PS (13 mm, range 4–110 mm) (p = 0.001). In patients with pNENs \ 20 mm (n = 27), the size difference (median RS 15 mm vs PS 12 mm) was also significant (p = 0.003). However, the only modality that significantly overestimated the PS was EUS (median RS 14 mm vs 11 mm; p = 0.0002). RS overestimated the PS in 21 patients (21 of 27 patients, 78%). Five of 11 patients (12%) with a Nf-pNEN and a RS [ 20 mm had in reality a PS \ 20 mm. MRI was the imaging technique that best correlated with PS in the total cohort (r = 0.8; p \ 0.0001), whereas EUS was the best correlating imaging tool in pNENs \ 20 mm (r = 0.5; p = 0.0001). Conclusion Preoperative imaging, especially EUS, frequently overestimates the size of MEN1-pNENs, especially those with a PS \ 20 mm. This should be considered when indicating surgery in MEN1 patients with small NfpNENs.

& V. Polenta [email protected] 1

Department of Visceral, Thoracic and Vascular Surgery, Philipps-University Marburg, Marburg, Germany

2

Division Endocrinology, Philipps-University Marburg, Marburg, Germany

3

Institute of Pathology, Philipps-University Marburg, Marburg, Germany

4

Department of Radiology, Philipps-University Marburg, Marburg, Germany

5

Department of General Surgery, Ospedali Riuniti Ancona, Conca 71, Torrette, 60020 Ancona, Italy

123

World J Surg

Introduction Multiple endocrine neoplasia Type 1 (MEN1) is an autosomal dominant inherited tumor syndrome characterized by the development of neuroendocrine tumors of different endocrine organs including the parathyroid glands, the pancreas and duodenum, the anterior pituitary gland and, less frequent, the adrenal glands, the thymus and the bronchii [1–5]. MEN1 is caused by line mutations in the Menin suppressor gene identified on chromosome 11q13 [5, 6]. The penetrance of the MEN1 syndrome is approximately 80–98% by the age of 50 years [7, 8]. According to more recent data, pancreatic neuroendocrine neoplasms (pNENs) will be diagnosed in up to 80% of MEN1 patients in prospective screening programs [9, 10]. One of the characteristics of MEN1-pNENs is that these occur multiple in the majority of patients and might be functioning or non-functioning, benign or malignant. In MEN1, non-functioning pNENs (Nf-pNENs) are most common with up to 80% [8, 10]. The prevalence of these lesions is steadily increasing as a result of earlier diagnosis based on genetic testing, standardized screening programs and more sensitive imaging studies, especially endoscopic ultrasound (EUS) [11, 12]. Among the functioning neoplasms, gastrinomas are the most common, representing up to 60% of the functioning MEN1-related pNENs. Insulinomas occur in 10–33% of patients with MEN1, whereas other functioning neoplasms such as VIPomas and glucagonomas are rarely observed [5–8]. The detection and adequate treatment of pNENs is important, since pNENs and thymic carcinoid are the most common cause of death in MEN1 [13]. However, the management of MEN1pNENs remains still controversial. It is consensus that insulinomas, glucagonomas and vipomas should be resected if there is no diffuse distant metastatic disease [7, 14, 15]. The timing and indication of surgery for MEN1-gastrinoma are highly controversial, ranging from medical treatment with just PPIs to surgery, if any imaginable pNEN reaches 2 cm in size or even at the time of biochemical diagnosis to prevent the development of metastatic disease. It has been reported that Nf-pNENs \ 2 cm in size rarely (\10%) develop metastases and NfpNENs \ 1 cm almost never develop metastases [12, 16]. Current practice guidelines recommend surgery in MEN1associated Nf-pNENs with a size [1 cm on imaging, whereas ENETS and NANETS guidelines recommend resection only if the sizes is 2 cm or more [14, 15]. During several operations for pNENs in MEN1 patients, the senior author had the macroscopic impression that the tumor size in the resected specimens was frequently smaller than determined by preoperative imaging. Since the radiological

123

size an important parameter to indicate surgery in MEN1associated gastrinoma and Nf-pNENs, the aim of this study was to determine the accuracy of the preoperatively determined radiological size compared to the pathological size of MEN1-pNENs.

Materials and methods After approval of the local ethics committee in 1997, all MEN1 patients treated at the Department of Surgery of the University Hospital Marburg were documented in a prospective database. The diagnosis of MEN1 was based on a careful personal and family history as well as the identification of a germ line mutation in the MEN1 gene. In 1997, an annual routine screening was established at our institution which includes: careful personal and family anamnesis, focused on symptoms caused by functional tumors, biochemical analysis and imaging techniques [9]. Imaging techniques included endoscopic ultrasound (EUS), computed tomography (CT), magnetic resonance imaging (MRI) and/or somatostatin receptor scintigraphy (SRS) until 2012. After 2012, instead of a somatostatin receptor scintigraphy a Galium68--DOTATOC-PET-CT was performed. All CT and MRI scans were judged by experienced radiologists with a long-standing experience in cross-sectional imaging of GEP-NENs within our ENETS center of excellence. According to the ENETS guidelines [14], computed tomography (CT) represented the most common initial diagnostic tool in our institution until 2009. Since 2010, MRI was the preferred imaging tool to reduce the radiation load. EUS, including the documentation of the number and sizes of pancreatic lesions, was performed by one single very experienced investigator (PHK) with more than 15 years EUS experience for pNENs using a Pentax FG 32 UA endoscope with a longitudinal 7.5 MHz sector array in combination with a Hitachi EUB 525 ultrasound computer. The examination of the pancreas followed a standardized protocol according to institutional standard operating procedures. In case a relevant lesion or an indeterminate lesion was identified, the lesion was described, measured in two dimensions and photograph recorded [17–21]. The diagnosis of ZES was based on clinical symptoms such as recurrent ulcera, an elevated fasting gastrin level ([125 pg/ml) and a pathological secretin-test, defined as an increase in serum gastrin concentration to [200 pg/ml, together with low pH in the stomach, and a positive immunohistochemistry in the tumor. The diagnosis of insulinoma was established by a symptomatic hypoglycemia (\45 mg/dl) with concomitant endogenous hyperinsulinism ([20 lU/ml) during a supervised fasting test and a positive immunohistochemistry for insulin of the tumor. PNENs were considered non-functioning (Nf), if

World J Surg

clinical signs of hormonal excess were absent and plasma hormone levels—despite pancreatic polypeptide—were within normal levels. In the present study, patients who underwent surgery between 2000 and 2014 for functioning and Nf-pNENs without distal metastases or vascular involvement, were included. Surgery was indicated in patients with a functioning pNENs after the exclusion of diffuse distant metastases by preoperative imaging. Preoperative EUS-guided fine-needle biopsy of pancreatic lesions was not routinely performed, since all pancreatic tumors were assumed to be pNENs in the cohort of MEN1 patients. Only a small proportion of MEN1 patients, who had small, but fast growing tumors within 1 year, underwent EUS-guided biopsy to exclude G2/G3 pNENs. Until 2010 patients with a Nf-pNENs were scheduled for pancreatic resection, if the tumor size extended 10 mm in preoperative imaging and diffuse liver metastases were excluded. Since 2011 Nf-pNENs were operated according to ENETS guidelines, if the largest tumor size reaches 20 mm [14]. In addition, we scheduled patients with NfpNENs with a size between 10 and 20 mm for surgery, if a more than 20% growth was observed during 12 months follow up and/or a CHES1 LOI was detected in the particular patient [22]. Depending of the tumor number and location, as determined by intraoperative ultrasonography (IOUS), standard pancreatic resections such as pancreaticoduodenectomy, distal pancreatectomy with or without spleen preservation or enucleation were performed. Parenchyma-preserving resections, such as enucleation, were preferred, if technically possible, for insulinomas and well encapsulated, small (\2 cm) Nf-pNENs without malignancy features [23]. The indication for reoperation was principally the same as for primary surgery. The specific reoperation performed was dependent on the pattern of disease recurrence as visualized by imaging studies. Tumors were classified according to the World Health Organization (WHO) System and malignancy was determined based on the existence of lymph node or distant metastases [24]. According to the Ki67 proliferation index, the tumors were classified as pNEN low grade (G1), intermediate grade (G2) and high grade (G3) [24]. Radiological tumor size (RS) was defined as the largest transverse diameter of the largest imaged pNEN by endoscopic ultrasound (EUS), magnetic resonance imaging (MRI) or computed tomography (CT). In addition, it was analyzed whether preoperative imaging revealed solitary or multiple pNENs. In the present study, somatostatin receptor scintigraphy (SRS) and Galium68--DOTATOC-PET-CT were not included because such a measurement was not available for all the patients and size measurement was not reliable. After surgical resection, specimens were immediately submitted to surgical pathology for frozen section to confirm diagnosis and for biobanking. Tumor

dimensions were determined by measurement of the fresh specimen using a metric ruler with millimeter demarcations. The largest recorded transverse tumor dimension from the surgical pathology report was used as pathological size (PS) for statistical analysis. For patients with multiple pNENs, the size of the largest lesion was included in the analysis. Statistical analysis: Distribution of continuous variables is reported as median and range. Categorical variables are presented as numbers and percentages. The paired Student’s t test was used to compare the median radiological tumor size and the median pathological tumor size. Statistical significance was defined as p less or equal than 0.05. A regression analysis was applied to study the linear association between radiographic and pathologic size. Statistical analyses were performed in SPSS 15.0 for Windows (software).

Results A total of 44 MEN1 patients were identified, including 25 men (57%) and 19 women (43%) with a median age of 37 (range 10–68) years at the time of diagnosis. PNENs were functioning in 18 patients (41%) and non-functioning in 26 patients (59%). Thirty-seven patients (84%) had multiple pNENs on imaging, while seven patients (16%) had single lesion. In fourteen years, 39 patients (89%) underwent a primary resection and 5 patients (11%) a pancreatic reoperation for pNENs. According to the tumor location and tumor type, 7 patients (16%) had a pancreaticoduodenectomy for ZES, 18 patients (41%) had distal pancreatectomy, 8 patients (18%) underwent a combined operation comprising spleen-preserving distal pancreatectomy with enucleation of pancreatic head tumors, 8 patients (18%) had only enucleations and 3 patients (7%) underwent completion pancreatectomy, respectively. Forty patients (91%) had pNEN G1, 2 patients (4%) pNEN G2 and 2 patients (4%) had multifocal G1 and G2 tumors. The pathological tumor stage according to the ENETS classification was stage I in 27 (61%) patients, stage II in 8 (18%) patients and stage III in 9 (20%) patients. The clinical and pathological characteristics of the entire cohort are summarized in Table 1. Forty patients (91%) had an EUS, 16 patients (36%) a CT scan and 18 (41%) patients underwent MRI. Table 2 describes the comparison between the histological and pathological diameter in the total cohort according to imaging procedure. The median radiological tumor size (RS), measured in all patients who had EUS, CT or MRI, was 20 mm (range 3–100 mm) compared to a median pathological tumor size (PS) of 13 mm (4–110 mm). Thus, the RS significantly overestimated the PS by 7 mm (p = 0.001). This result held true in

123

World J Surg

the subgroup of patients who had a primary resection (p = 0.007) and tended to be significant in the 5 patients who underwent a second resection (p = 0.062). The median RS was 19 mm (range 4–50 mm) on EUS, 18 mm (range 5–36 mm) on CT and 18 mm (range 6-100 mm) on MRI compared to the median PS of 12 mm (4–50 mm) for patients who had EUS, 14 mm (4–40 mm) for patients who had CT and 15 mm (4–110 mm) for patients who underwent MRI, respectively. Thus, EUS, CT and MRI overestimated the pathological size by median 7, 4 and 3 mm, respectively. However, this difference was only significant for the RS measured by EUS (p = 0.013). In the subgroup of patients (n = 27) with a small pNEN (\20 mm) (Table 3), the median RS of all imaging methods was, with 15 mm (3–36 mm), significantly larger than the median PS with 12 mm (4–20 mm); (p = 0.003). Considering the RS for each imaging modality, the median RS was 14 mm (4–26 mm) for EUS, 15 mm (5–36 mm) for abdominal CT and 15 mm (6–22 mm) for MRI compared to the median pathological tumor size (PS) of 11 mm (4–20 mm) for patients who underwent EUS, 12 mm (5–20 mm) for patients who had CT and 14 mm (4–20 mm) who had MRI, respectively. The size difference was only significant for the measurement obtained by EUS (p = 0.0002). In patients with a tumor \20 mm RS overestimated the PS in 21 patients (78%) and underestimated the PS in 3 patients (11%). Eleven patients of 27 with a NfpNEN had a preoperative RS [ 20 mm, but definitive histopathology revealed a PS \ 20 mm in 5 of these 11 patients. The graphics in panel A-B-C (Fig. 1) reflect the linear regression of maximum tumor diameter at pathology and at imaging for EUS, CT and MRI. Overall, MRI determined the RS better than the other procedures in the total cohort (r = 0.8; p \ 0.0001), but not in the subgroup with a lesion\20 mm (r = 0.4). For these tumors the most accurate was EUS, although the r value was only 0.5. Table 4 describes the sensitivity of imaging in MEN1pNENs. EUS represents the imaging tool that best identified a pNENs, especially in the presence of multiple small tumors.

Discussion This is the first study that evaluated the accuracy of radiological imaging with regard of pNEN size compared to pathological size in MEN1. The presented data clearly demonstrate that preoperative imaging (RS) frequently tends to overestimate the true pathological size (PS), especially in pNENs with a diameter \20 mm. Similar results have been reported in hepatocellular carcinoma and renal cell carcinoma which are also very well-vascularized tumors. For renal cell carcinoma, it appears that CT usually

123

Table 1 Clinical and pathological characteristics of the entire cohort (n = 44) Variable

n (%)

Gender Male

25 (57)

Female

19 (43)

Age (median years)

37 (10; 68)

Tumor localization Head

7 (16)

Body/tail Multifocal

18 (41) 19 (43)

Non-functioning

26 (59)

Functioning

18 (41)

Type of surgical resection Pancreaticoduodenectomy

7 (16)

Distal pancreatectomy alone

18 (41)

Completion pancreatectomy

3 (7)

Enucleation

8 (18)

Distal pancreatectomy ? enucleation pancreatic head

8 (18)

Surgery Primary resection

39 (89)

Reoperation

5 (11)

Tumor grading NEN-G1

40 (91)

NEN-G2 NEC-G3

2 (4) 0

NEN—G1 ? G2

2 (4)

Stage I

27 (61)

II

8 (18)

III

9 (20)

IV

0

overestimates the true pathologic size, especially for smaller tumors [25, 26]. Also for hepatocellular carcinoma, in most instances (81%), imaging by CT or MRI overestimates true pathologic diameter [27]. The correlation between radiological and pathological size in MEN1-associated pNENs is very important, since the indication for surgery depends on the pNEN size according to current ENETS [14] and expert guidelines [12, 13], especially for Nf-pNENs and MEN1-associated gastrinomas. The pNEN size is positively correlated with the metastatic potential in MEN1-pNENs [28]. Triponez et al. who reported the French experience, observed a correlation between size of the tumors and their malignancy: a diameter of 1,5-2 cm is considered a negative factor, which marks the malignancy risk [12, 29]. The experience of Norton et al. has reported that the risk of hepatic metastases in patients with MEN1-gastrinomas increases with tumor size: 25–40% of patients with

World J Surg Table 2 Comparison of radiological and pathological size in the total cohort (n = 44) Imaging technique

N (%)

Radiological size (mm), median (range)

Pathological size (mm), median (range)

p

r

p

All imaging methods

44 (100)

20 (3–100)

13 (4–110)

0.001

0.8056 \0.0001

EUS

40 (91)

19 (4–50)

12 (4–50)

0.013

0.5873 \0.0001

CT

16 (36)

18 (5–36)

14 (4–40)

0.456

0.187

MRI

18 (41)

18 (6–100)

15 (4–110)

0.893

0.8307 \0.0001

0.09

EUS endoscopic ultrasonography, CT computed tomography, MRI magnetic resonance imaging

Table 3 Comparison of radiological and pathological size in patients with pNEN \ 20 mm Imaging technique

N (%)

Radiological size (mm), median (range)

Pathological size (mm), median (range)

p

r

p

All imaging methods

27 (100)

15 (3–36)

12 (4–20)

0.003

0.0898

0.1288

EUS

25 (93)

14 (4–26)

11 (4–20)

0.0002

0.4725

0.0001

CT

13 (48)

15 (5–36)

12 (5–20)

0.217

0.1036

0.3643

MRI

10 (37)

15 (6–22)

14 (4–20)

0.280

0.3931

0.0524

EUS endoscopic ultrasonography, CT computed tomography, MRI magnetic resonance imaging

pancreatic NET larger than 4 cm develop hepatic metastases, and 50–70% of MEN1 patients with tumors 2–3 cm in size have lymph node metastases [30, 31]. About NfpNENs, other authors have been previously demonstrated that 27% of MEN1 patients with Nf-pNENs between 2.1 and 3 cm developed metastases compared to 11% of patients with Nf-pNENS \ 2 cm [12, 32, 33]. It has been shown in previous retrospective studies that surgery for NfpNENs \ 2 cm is not beneficial compared to close surveillance [16, 30, 34]. For MEN1-ZES, an imaginable pNEN [ 2 cm seems to be a good surrogate marker to indicate surgery, since the fast majority will not develop distant metastases although they will be rarely cured biochemically [35, 36]. The optimal operative approach for these patients should be based on a careful balance of the potential oncologic benefits against the perioperative risks, long-term pancreatic organ function and health-related quality of life (QQL) [37]. Actually, no evidence-based consensus exists concerning the best timing for surgery. The MEN1 clinical practice guidelines recommend an aggressive approach considering surgery for a lesion [1 cm [7], whereas ENETS und NANETS guidelines recommend surgery for lesions [2 cm [14, 15]. Here, we demonstrate that the preoperative imaging techniques frequently overestimate the final pathological size in MEN1 patients that underwent EUS, CT or MRI. However, analysis of distinct imaging techniques revealed, that this result is only statistically significant and clinically relevant for EUS, especially in the subgroup of patients who had a small tumor (\20 mm). In 78% of these patients, the RS overestimated the PS and 46% of the patients with a Nf-

pNEN [ 20 mm on imaging, indeed had a tumors \20 mm. Thus, these patients would no have required surgery according to actual ENETS and NANETS guidelines [14, 15]. Although EUS tended to overestimate the tumor diameter in the subgroup of patients with pNENs \ 20 mm, it represented still the best correlated imaging to estimate true pathologic size with a r value of 0.5 (p = 0.0001) compared to CT (r value of 0.1, p = 0.36; MRI r value of 0.4, p = 0.05). CT and MRI are frequently inefficient for the detection of small pNENs (\1 cm) as shown in the present study. This in line with several studies in the literature that have demonstrated EUS as the most sensitive tool for the detection of small pNENs [17, 18, 38–40]. This discrepancy of EUS being the best imaging modality to identify small pNENs on one hand, but frequently overestimates the PS, especially of small tumors, on the other hand, remains unclear. One possible explanation might be that the resection of the well-vascularized pNENs interrupts the blood circulation. This might lead to a reduced tumor volume after resection resulting in a reduced tumor diameter on pathological examination. This might be also the case for cross-sectional imaging with CT and MRI. Another explanation for the latter two imaging modalities might be that helical DCT and MRI processes potentially yield tumor motion measurement errors, especially by breathing [41, 42]. This led to tumor size overestimation in almost 40% of patients with abdominal cancers in one previous technical study [41]. In EUS, it can be sometimes difficult to exactly determine the size, if the pNEN is isoechogenic and not well encapsulated even for a very experienced examiner.

123

World J Surg

A

Pathological Size vs EUS Size: 40 patients

60 50

Pathological Size

Fig. 1 Linear regression plots of maximum tumor diameter (mm) at pathology and at imaging for EUS, CT and MRI. Panel A: EUS (n = 40); Panel B: CT (n = 16); Panel C: MRI (n = 18)

40 30 20 10 0

0

10

20

30

40

50

60

EUS Size

Pathological size

B

Pathological Size vs CT size: 16 patients

45 40 35 30 25 20 15 10 5 0

0

5

10

15

20

25

30

35

40

CT

C

Pathological Size vs MRI Size: 18 patients 120

Pathological Size

100 80 60 40 20 0

0

20

40

60

80

100

120

MRI

The present study has some limitations. First, its retrospective design and the limited number of patients, but MEN1 is a rare disease. Secondly, not all patients underwent the same three imaging procedures, which induces some bias. Third, EUS is more operator dependent than CT or MRI imaging. Even though the procedure was performed by one very experienced endoscopist and the procedure was performed with the same endoscope according

123

to the same standard, a systematic error or bias cannot be completely excluded. Based on these data, our group changed its aggressive policy to indicate surgery in Nf-pNENs based on a certain size. Nowadays, we are very restrictive to indicate surgery in Nf-pNENs between 1 and 2 cm. Even the proposed cutoff [2 cm to indicate surgery will now be considered

World J Surg Table 4 Sensitivity of imaging in MEN1-pNENs Imaging EUS (N = 40) CT (N = 16) MRI (N = 18)

Solitary tumors detecteda 7

Multiple tumors detecteda

pNENs \ 10 mm detecteda

pNENs detected only by this imaging technique

33

20

14

11

5

3

2

9

9

7

1

a

As determined by pathological analysis

Fig. 2 Suggested algorithm for the diagnostic workup of MEN1-NF-pNENs based on imaging size. Asterisk indicates patients [65 years and patients with general risk factors surgery may be postponed until a size of 23 mm; #-CHES1-LOI as determine by the location of the MEN1 mutation [22]; LKmets—lymph node metastases

more liberally, since EUS overestimate the size of about 3 mm (Fig. 2).

Conclusion Preoperative imaging, particularly EUS, frequently overestimates the size of MEN1-pNENs, especially those with a pathologic size \20 mm. Such a potential overestimation has to be considered when indicating surgery in MEN1 patients with small Nf-pNENs detected during screening programs. Acknowledgements We thank all patients who participated in our screening program. Compliance with ethical standards Conflict of interest The authors declare that they have no conflict of interest.

References 1. Brandi ML (2000) Multiple endocrine neoplasia type 1. Rev Endocr Metab Disord 1(4):275–282 2. Trump D, Farren B, Wooding C et al (1996) Clinical studies of multiple endocrine neoplasia type 1 (MEN1). QJM 89(9):653–669 3. Marx S, Spiegel AM, Skarulis MC et al (1998) Multiple endocrine neoplasia type 1: clinical and genetic topics. Ann Intern Med 129(6):484–494 4. Carney JA (2005) Familial multiple endocrine neoplasia: the first 100 years. Am J Sur Pathol 29(2):254–274 5. Thakker RV (2014) Multiple endocrine neoplasia type 1 (MEN1) and type 4 (MEN4). Mol Cell Endocrinol 386(1–2):2–15 6. Chandrasekharappa SC, Guru SC, Manickam P et al (1997) Positional cloning of the gene for multiple endocrine neoplasia type 1. Science 276(5311):404–407 7. Thakker RV, Newey PJ, Walls GV et al (2012) Clinical practice guidelines for multiple endocrine neoplasia type 1 (MEN1). J Clin Endocrinol Metab 97(9):2990–3011 8. Tonelli F, Giudici F, Giusti F et al (2012) Gastroenteropancreatic neuroendocrine tumors in multiple endocrine neoplasia type 1. Cancers 4(2):504–522

123

World J Surg 9. Waldmann J, Fendrich V, Habbe N et al (2009) Screening of patients with endocrine neoplasia type 1 (MEN-1): a critical analysis of its value. World J Surg 33(6):1208–1218. doi:10. 1007/s00268-009-9983-8 10. Thomas-Marques L, Murat A, Delemer B et al (2006) Prospective endoscopic ultrasonographic evaluation of the frequency of non functioning pancreaticoduodenal endocrine tumors in patients with multiple endocrine neoplasia type 1. Am J Gastroenterol 101:266–273 11. Carty SE, Helm AK, Amico JA et al (1998) The variable penetrance and spectrum of manifestations of multiple endocrine neoplasia type 1. Surgery 124(6):1106–1113 12. Triponez F, Dosseh D, Goudet P et al (2006) Epidemiology data on 108 MEN 1 patients from the GTE with isolated nonfunctioning tumors of the pancreas. Ann Surg 243(2):265–272 13. Goudet P, Murat A, Binquet C et al (2010) Risk factors and causes of death in MEN1 disease. A GTE (Groupe d’Etude des Tumeurs Endocrines) cohort study among 758 patients. World J Surg 34(2):249–255. doi:10.1007/s00268-009-0290-1 14. Falconi M, Eriksson B, Kaltsas G et al (2016) ENETS consensus guidelines update for the management of patients with functional pancreatic neuroendocrine tumors and non-functional pancreatic neuroendocrine tumors. Neuroendocrinology 103(2):153–171 15. Vinik AI, Woltering EA, Warner RR et al (2010) NANETS consensus guidelines for the diagnosis of neuroendocrine tumor. Pancreas 39(6):713–734 16. Partelli S, Tamburrino D, Lopez C et al (2016) Active surveillance versus surgery of nonfunctioning pancreatic neuroendocrine neoplasms B2 cm in MEN1 patients. Neuroendocrinology 103(6):779–786 17. Baur AD, Pavel M, Prasad V et al (2016) Diagnostic imaging of pancreatic neuroendocrine neoplasms (pNEN): tumor detection, staging, prognosis, and response to treatment. Acta Radiol Open 57(3):260–270 18. Kann P, Bittinger F, Engelbach M et al (2001) Endosonography of insulin-secreting and clinically non-functioning neuroendocrine tumors of the pancreas: criteria for benignancy and malignancy. Eur J Med Res 6(9):385–390 19. Kann PH, Wirkus B, Keth A et al (2003) Pitfalls in endosonographic imaging of suspected insulinomas: pancreatic nodules of unknown dignity. Eur J Endocrinol 148(5):531–534 20. Kann PH, Balakina E, Ivan D et al (2006) Natural course of small, asymptomatic neuroendocrine pancreatic tumors in multiple endocrine neoplasia type 1: an endoscopic ultrasound imaging study. Endocr Relat Cancer 13(4):1195–1202 21. Kann PH, Kann B, Fassbender WJ et al (2006) Small neuroendocrine pancreatic tumors in multiple endocrine neoplasia type 1 (MEN1): least significant change of tumor diameter as determined by endoscopic ultrasound (EUS) imaging. Exp Clin Endocrinol Diabetes 114(7):361–365 22. Bartsch DK, Slater EP, Albers MB et al (2014) Higher risk of aggressive pancreatic neuroendocrine tumors in MEN1 patients with MEN1 mutations affecting the CHES1 interacting MENIN domain. J Clin Endocrinol Metab 99(11):2387–2391 23. Crippa S, Bassi C, Salvia R et al (2007) Enucleation of pancreatic neoplasms. Br J Surg 94(10):1254–1259 24. Klo¨ppel G, Perren A, Heitz PU (2004) The gastroenteropancreatic neuroendocrine cell system and its tumors: the WHO classification. Ann N Y Acad Sci 1014:13–27 25. Jeffery NN, Douek N, Guo DY et al (2011) Discrepancy between radiological and pathological size of renal masses. BMC Urol 11:2

123

26. Irani J, Humbert M, Lecocq B et al (2001) Renal tumor size: comparison between computed tomography and surgical measurements. Eur Urol 39(3):300–303 27. Kelsey CR, Schefter T, Nash SR et al (2005) Retrospective clinicopathologic correlation of gross tumor size of hepatocellular carcinoma: implications for stereotactic body radiotherapy. Am J Clin Oncol 28(6):576–580 28. Doherty GM, Olson JA, Frisella MM et al (1998) Lethality of multiple endocrine neoplasia type 1. World J Surg 22(6):581–586. doi:10.1007/s002689900438 29. Triponez F, Goudet P, Dosseh D et al (2006) Is surgery beneficial for MEN 1 patients with small (\2 cm), non-functioning pancreaticoduodenal endocrine tumor? An analysis of 65 patients from GTE. World J Surg 30(5):654–662. doi:10.1007/s00268005-0354-9 30. Norton JA, Fraker DL, Alexander HR et al (1999) Surgery to cure the Zollinger–Ellison syndrome. N Engl J Med 341(9):635–644 31. Norton JA (2005) Surgical treatment and prognosis of gastrinoma. Best Pract Res Clin Gastroenterol 19(5):799–805 32. Imamura M, Takahashi K (1993) Use of selective secretin injection test to guide surgery in patients with Zollinger Ellison syndrome. World J Surg 17(4):433–438. doi:10.1007/ BF01655100 33. Skogseid B, Eriksson B, Lundgvist G et al (1991) Multiple endocrine neoplasia type 1: a 10 years prospective screening study in four kindreds. J Clin Endocrinol Metab 73(2):281–287 34. Bartsch DK, Langer P, Wild A et al (2000) Pancreaticoduodenal endocrine tumors in multiple endocrine neoplasia type 1: surgery or surveillance? Surgery 128(6):958–966 35. Morrow EH, Norton JA (2009) Surgical management of Zollinger-Ellison syndrome; state of the art. Surg Clin North Am 89(5):1091–1103 36. Lopez CL, Albers MB, Bollmann C et al (2016) Minimally invasive versus open pancreatic surgery in patients with multiple endocrine neoplasia type 1. World J Surg 40(7):1729–1736. doi:10.1007/s00268-016-3456-7 37. You YN, Thompson GB, Young WF Jr et al (2007) Pancreatoduodenal surgery in patients with multiple endocrine neoplasia type 1: operative outcomes, long-termfunction, and quality of life. Surgery 142(6):829–836 38. Gauger PG, Scheiman JM, Wamsteker EJ et al (2003) Role of endoscopic ultrasonography in screening and treatment of pancreatic endocrine tumours in asymptomatic patients with multiple endocrine neoplasia type 1. Br J Surg 90(6):748–754 39. Langer P, Kann PH, Fendrich V et al (2004) Prospective evaluation of imaging procedures for the detection of pancreaticoduodenal endocrine tumours (PETs) in patients with multiple endocrine neoplasia type 1 (MEN 1). World J Surg 28(12):1317–1322. doi:10.1007/s00268-004-7642-7 40. Van Asselt SJ, Brouwers AH, van Dullemen HM et al (2015) EUS is superior for detection of pancreatic lesions compared with standard imaging in patients with multiple endocrine neoplasia type 1. Gastrointest Endosc 81(1):159–167 41. Dou TH, Thomas DH, O’Connel D et al (2015) Technical note: simulation of 4DCT tumor motion measurement errors. Med Phys 42(10):6084–6089 42. Haraldsdo`ttir KH, Jo`nsson P, Halldo`rsdo`ttir AB et al (2017) Tumor size of invasive breast cancer on magnetic resonance imaging and conventional imaging (mammogram/ultrasound): comparison with pathological size and clinical implications. Scand J Surg 106(1):68–73