Cystic Pancreatic Neuroendocrine Tumors (Cpnets): a Systematic Review and Meta-Analysis of Case Series Luis Hurtado-Pardo1, Javier A

REVIEW

Cystic pancreatic neuroendocrine tumors (cPNETs): a systematic review and meta-analysis of case series Luis Hurtado-Pardo1, Javier A. Cienfuegos1-3, Miguel Ruiz-Canela2,3, Pablo Panadero4, Alberto Benito5 and José Luis Hernández-Lizoain1 1Department of General Surgery. Clínica Universidad de Navarra. Pamplona, Spain. 2Department of Preventive Medicine and Public Health. Universidad de Navarra. Pamplona, Spain. 3Institute of Health Research of Navarra (IdisNA). Pamplona, Spain. 4Department of Pathology, and 5Department of Radiology. Clínica Universidad de Navarra. Pamplona, Spain

ABSTRACT Establishing a differential diagnosis with regard to other cystic and solid lesions of the pancreas may be difficult Cystic pancreatic neuroendocrine tumors represent 13% of all as in many cases cPNETs are diagnosed incidentally and neuroendocrine tumors. The aim of this study is to analyze the phenotype and biologic exhibit great clinical variability: functional, non-function- behavior of resected cystic neuroendocrine tumors. al, sporadic or hereditary (3-5). A systematic review and meta-analysis were conducted until The aim of the present study is to carry out a systematic September 2016 using a search in Medline, Scopus, and EMBASE review and meta-analysis of resected cystic pancreatic neu- with the terms “cystic pancreatic endocrine neoplasm”, “cystic islets roendocrine tumors reported in the literature, including 5 tumors” and “cystic islets neoplasms”. From the 795 citations recovered 80 studies reporting on 431 cases from our own center, with the objective of obtaining patients were selected. 87.1% (n = 387) were sporadic tumors and the most exhaustive scientific knowledge on their clinical, 10.3% (n = 40) corresponded to multiple endocrine neoplasia type pathologic and prognostic characteristics. 1. Were diagnosed incidentally 44.6% (n = 135). Cytology was found to have a sensitivity of 78.5%. Were non-functional tumors 85% (n = 338), and among the functional tumors, insulinoma was the most frequent. According to MATERIAL AND METHODS the European Neuroendocrine Tumor Society staging, 87.8% were limited to the pancreas (I-IIb), and 12.2% were advanced (III-IV). The systematic review was carried out following the recommen- Disease-free survival at 5 years in stages (I-IIIa) and (IIIb-IV) was 91.5% dations reported in the “Preferred Reporting Items for Systematic and 54.2%, respectively; and was significantly lower (p = 0.0001) in Reviews and Meta-Analyses” (PRISMA) adapted to surgical series functional tumors. In patients with multiple endocrine neoplasia there was a higher incidence of functional (62.5%) and multifocal (28.1%) and cases (6,7). tumors. Disease-free survival at 5 and 10 years was 60%. Cystic pancreatic neuroendocrine tumor was defined as any neu- Cystic pancreatic neuroendocrine tumors exhibit phenotypical roendocrine tumor which had a cystic pattern (purely cystic or mixed characteristics which are different to those of solid neuroendocrine cystic-solid lesions) in the macroscopic study or the imaging studies tumors. and confirmed by histology. – Inclusion criteria: all published cases of cystic pancreatic Key words: Cystic pancreatic neuroendocrine tumors. Multiple neuroendocrine tumors which were treated with surgery and endocrine neoplasia type 1. Cystic tumors. Oncologic outcomes. Evidence base medicine. with histologic confirmation were identified. Non-resected tumors or those where histologic confirmation was lacking were excluded. nd INTRODUCTION – Methods use to search for and identify studies: up to 22 Sep- tember 2016, the following databases were consulted: Medline, Cystic pancreatic neuroendocrine tumors (cPNETs) Scopus, and EMBASE. Studies in English, German French and Spanish were included. A search of the “grey” literature was account for between 13% and 17% of pancreatic neuro- also undertaken with the inclusion of cases reported at con- endocrine tumors (PNETs) and 9-10% of resected cystic ferences, letters to the editor or chapters in books. The search tumors (1,2). terms used were “cystic pancreatic endocrine neoplasms”, Due to the widespread use of diagnostic methods with “cystic islets tumors”, and “cystic islets neoplasms”. greater sensitivity and specificity the incidence of such The articles identified were recorded in an Excel® database tumors has risen in recent years and has led to the publi- and duplicates were eliminated. In the database, the titles and cation of cases and clinical series. abstracts of the articles were independently reviewed by two

Received: 08-05-2017 Accepted: 07-08-2017 Hurtado-Pardo L, Cienfuegos JA, Ruiz-Canela M, Panadero P, Benito A, Hernández-Lizoain JL. Cystic pancreatic neuroendocrine tumors (cPNETs): Correspondence: Javier A. Cienfuegos. Department of General Surgery. a systematic review and meta-analysis of case series. Rev Esp Enferm Dig Clínica Universidad de Navarra. University of Navarra. Av. Pío XII, 36. 2016;109(11):778-787. 31008 Pamplona, Spain DOI: 10.17235/reed.2017.5044/2017 e-mail: [email protected] 2017, Vol. 109, N.º 11 CYSTIC PANCREATIC NEUROENDOCRINE TUMORS (cPNETs): A SYSTEMATIC REVIEW 779 AND META-ANALYSIS OF CASE SERIES

authors (LH, JAC) in order to assess the validity and quality (SD = 16.1). Male patients were significantly older (57.1 of the case series. The full text of all the articles selected was vs. 49.5; p = 0.001). printed to ensure the inclusion criteria had been met. When Genetic characteristics were reported in 387 patients: doubts arose, the possible reasons for excluding the article 337 patients (87.1%) were sporadic, 40 (10.3%) had MEN- were discussed by the team, and a final decision was taken 1 (multiple endocrine neoplasis type 1), 7 (1.2%) had Von by consensus. If multiple case series were published from the Hippel Lindau syndrome and 3 tuberous sclerosis (0.8%). same center with overlapping study periods, the publication Details of the clinical presentation were obtained in 303 with the largest number of cases was selected. patients, of whom 135 (44.6%) were asymptomatic and – Data extraction: data were extracted using a predefined pro- thus corresponded to “incidentalomas”, 106 (35%) pre- tocol which included sex, age, associated genetic disorder, sented with abdominal pain, 16 (15.3%) with an abdominal symptoms, presentation type, diagnostic methods, fine needle mass, 14 (4.6%) with weight loss, 10 (3.3%) diabetes on punction-aspiration (FNPA) cytology, size and location of the tumor, type of surgery, post-operative complications, immuno- diagnosis and 8 (2.6%) pancreatitis. histochemical and pathologic characteristics, stage according Data on the functionality of the tumors were available to ENETS (European Neuroendocrine Tumor Society) (8), in 339 cases. In 287 cases (85%) they were non-function- Ki-67 grade as defined by the WHO (World Health Organiza- al while in 52 (15.3%) they were functional: 18 insuli- tion) (9), survival and time to tumor recurrence. All data were nomas (5.3%), 12 glucagonomas (3.5%), 10 gastrino- extracted in duplicate to ensure accuracy. mas (3%), 7 somatostitomas (2.1%), 3 ACTH-secreting – Data synthesis and analysis: The continuous variables were tumors (ACTHoma) (0.9%), 1 vasoactive intestinal pep- summarized with means, ranges and standard deviations. In tide-producing tumor (VIPoma) (0.3%) and 1 growth hor- the series of patients where individual patient data were not mone-producing tumor (GHoma) (0.3%). provided, these data were not included in the analysis. Categor- In 196 patients a preoperative diagnosis was made ical variables were analyzed using Pearson’s χ2 test or Fisher’s using FNPA leading to a diagnosis of endocrine tumor in exact test. To compare continuous variables from independent samples the Student´s t-test was used or its nonparametric equivalent, the Mann-Whitney U test. The two-sided level of significance was set at p < 0.05. 384 studies identified Kaplan-Meier curves were plotted to assess survival, with a through Scopus and significance level of p < 0.05, and using the Log-rank test to Medline search analyze differences in survival depending on different characteristics. 411 studies identified All data were analyzed using Stata 12 software (StataCorp. through EMBASE Stata Statistical Software: Version 12.1. College Station, TX: search

Stata Corp LP). IDENTIFICATION TOTAL: 795 studies Duplicated RESULTS excluded Abstract not A total of 795 citations were identified. Once duplicate appropiate and unrelated articles had been excluded, 96 studies were 96 records emerged selected, from which another 5 were discarded as they did

SCREENING Studies excluded not deal with cystic PNETs. Of the remaining 91 articles, a (n = 5) further 11 were excluded: 6 articles because they reported on Non-cystic tumor the same series included in another article, 3 cases because they provided no data and 2 cases in which pathologic con- 91 articles found firmation of the tumor was not performed. Finally, 80 studies eligible and full-text Studies excluded were included, 67 isolated case reports (10-76) and 13 case reviewed (n = 11): Same serie (n = 6)

reviews (1,2,77-87). Of these 80 studies, the full text was not ELIGIBILITY available in 15, and so the relevant information was taken Without original data (n = 3) from the abstract. The 80 studies covered a total of 431 cases Without pathologic of cystic pancreatic neuroendocrine tumors. Five cases who 80 articles included in confirmation underwent surgery in our center were also included to make the review (n = 2) a final total on 436 cPNET cases. Figure 1 shows the flow 431 cases diagram of the articles included and reasons for exclusion as set out in the PRISMA guidelines (6). INCLUDED 5 own cases included The demographic and clinical characteristics of the 436 patients are shown in table 1. Of these patients 194 Fig. 1. Flow diagram for the inclusion of cases of cPNETs in the systematic (50.9%) were male, and the mean age was 53.2 years review following the PRISMA recommendations (6,7).

Rev Esp Enferm Dig 2017;109(11):778-787 780 L. HURTADO-PARDO ET AL. Rev Esp Enferm Dig

Table 1. Demographics and clinical characteristics Table 1 (Cont.). Demographics and clinical characteristics of cPNET tumor of cPNET tumor Variable n (%) Variable n (%) Gender Hemorraghe 55 (29.7) – Male 194 (50,92) Type of surgery – Female 187 (49,08) – Enucleation 28 (12.67) Age* mean (SD) 53.2 (16.12) – Whipple 31 (14.03) Genetic disorder – Corporocaudal pancreatectomy 20 (9.05) – Sporadic 337 (87.1) – Central pancreatectomy 7 (3.17) – MEN-1 40 (10.34) – Distal pancreatectomy 127 (57.47) – Von Hippel Lindau 7 (1.81) – Total pancreatectomy 7 (3.17) – Tuberous sclerosis 3 (0,78) – Marsupialization 1 (0.45) Clinical symptoms Ki-67 index – Asymptomatic 135 (44.55) – Low 186 (85.32) – Abdominal pain 106 (34.98) – Intermediate 32 (14.68) – Abdominal mass 16 (5.28) – High 0 (0) – Weight loss 14 (4.62) Stage (ENETS) – Diabetes 10 (3.3) – I 69 (25.7) – Pancreatitis 8 (2.63) – IIa 98 (36.4) – Jaundice 5 (1.65) – IIb 69 (25.7) – GI bleeding 3 (0.99) – IIIa 2 (0.7) – Previous pancreatitis 3 (0.99) – IIIb 18 (6.7) – Nausea/vomiting 3 (0.99) – IV 13 (4.8) Incidental diagnosis† 135 (44.55) Follow-up, months║, mean (SD) 42.49 (30.25) Endocrine symptoms Recurrence¶ 15 (8.02) – Non-functioning 287 (84.66) – Locoregional 5 (2.8) – Insulinoma 18 (5.31) – Distant 4 (2.1) – Glucagonoma 12 (3.54) – Non-specified 6 (3.2) – Gastrinoma 10 (2.95) Death** 5 (2.67) – Somatostitoma 7 (2.06) *Age: n = 201. †Incidental diagnosis: n = 303. ‡Citology-FNPA: n = 196. §Diam- eter: n = 193. ║Follow-up in months: n = 185. ¶Recurrence: n = 187. **Death: – ACTHoma 3 (0.88) n = 187. – Vipoma 1 (0.29) – GHRHoma 1 (0.29) 154 patients (78.6%) and inconclusive results in 20 cases Citology-FNA‡ (10.2%). – Neuroendocrine 154 (78.57) Tumor location was reported in 374 patients. Approx- – Non-concluyent 20 (10.2) imately half of the tumors, 184 (49.2%) were located in Location the tail, 87 (23.3%) in the head, 77 (20.6%) in the body, 7 (1.9%) in the neck and 6 (1.6%) in the uncinate process. – Head 100 (26.73) Thirteen cases (3.5%) were multifocal tumors. The maxi- – Body 77 (20.59) mum diameter reported in 193 patients was 43.8 mm (SD – Tail 184 (49.2) = 36.5). The type of surgical intervention was reported in 221 – Multiple 13 (3.48) patients: in 28 cases (12.7%) the procedure performed was Diameter (mm)§ mean (SD) 43.76 (36.48) enucleation, in 31 cases (14%) Whipple’s procedure, in Necrosis 22 (13.5) 20 cases (9.1%) corporocaudal pancreatectomy, in 7 cases

(Continue in the next column) (3.2%) central pancreatectomy, in 127 cases (57.5%) distal pancreatectomy, in 7 cases (3.2%) total pancreatectomy

Rev Esp Enferm Dig 2017;109(11):778-787 2017, Vol. 109, N.º 11 CYSTIC PANCREATIC NEUROENDOCRINE TUMORS (cPNETs): A SYSTEMATIC REVIEW 781 AND META-ANALYSIS OF CASE SERIES and in 1 case marsupialization. In 18 cases the procedures were performed laparoscopically. Information on ENETS staging was available for 269 patients: 69 (25.7%) were stage I, 98 (36.4%) stage IIa, 69 (25.7%) stage IIb, 2 (0.7%) stage IIIa, 18 (6.7%) stage IIIb and 13 (4.8%) stage IV. Lymph node involvement was reported in 295 patients with 268 cases (90.8%) with N0 and the remaining 27 (9.2%) with N1. Synchronous metastases were found in 13 patients (4.6%) of the 286 for whom this information was given. The mitotic index as defined by Ki-67 was reported in 218 of the tumors and was low in 186 (85.3%) and intermediate in 32 (14.9%). Twenty-two foci of necrosis (n = 163) and 55 foci of hemorrhage (n = 185) were identified. Follow-up was reported in 187 cases with a mean dura- tion of 42.5 months (SD = 30.3) and a range of 1 to 168 months. Out of these 187 cases, there were 15 cases of relapse (8%): 5 (2.8%) were locoregional, 4 (2.1%) distant and in 6 cases (3.2%) no location was given. Overall survival (OS) could be obtained for 185 cases, and 4 deaths were recorded (2.2%). Overall survival as calculated by ENETS staging in stages I and II at 5 and 10 years was 100%. In stage III, OS at 1 and 5 years was 100% and 85.7%, respectively. In stage IV, OS at 1 and 5 years was 100% and 75% respectively (p = 0.05). Disease-free survival rates (DFS) at 1 and 5 years in stages I-IIIa (without lymph node involvement) were 98.5%, 91.5% and 91.5% respectively while in stages IIIb-IV (with lymph node involvement) DFS at 1, 5 and 6 years was 100%, 54.2% y 54.2% respectively (p = 0.001) Fig. 2. Kaplan Meier curve for overall survival (A) and disease-free survival (Fig. 2). (B) following the ENETS classification for cPNET tumors.

Analyses by specific subgroup Comparison of both groups with regard to ENETS staging revealed that 41 cases (26.3%) were in stage I, 95 cases Functional vs. non-functional tumors (60.9%) in stage II, 14 cases (9%) in stage III and 6 cases (3,8%) in stage IV in the non-functional group, whereas in Table 2 summarizes the demographic and clinical char- the functional group there were 5 cases (15%) in stage I, acteristics of the 339 cases for whom the degree of func- 17 cases (51%) in stage II, 4 cases (12%) in stage III and tionality was reported. We found no significant differences 7 cases (21%) in stage IV (p = 0.007). regarding mean age at presentation between the two tumor Overall survival was similar in both groups (p = 0.68). types (49.2 vs. 50.3) (p = 0.896). In contrast, when we compared DFS in the non-func- Of the 35 functional tumors, 12 (34.3%) corresponded tional and functional cPNET, we found that patients with to patients with MEN-1, while in the group of 262 patients non-functional tumors had a significantly greater survival with non-functional tumors 20 were identified as having (p = 0.0001) than those with functional tumors. MEN-1 (7.7%) (p = 0.001). When location was compared, a greater proportion of multifocal tumors was found in the functional tumor group MEN-1 patients as compared to the non-functional group: 6 multifocal cases (17.6%) in the functional group (n = 34) as opposed to Table 3 summarizes the characteristics of the 40 cases 7 multifocal cases (2.7%) in the non-functional groups (n of hereditary MEN-1. The mean age of the 27 patients with = 260) (p = 0.016). No significant differences were found data was 41.7 years (SD=15). As for the degree of func- regarding size: in the non-functional tumor group (n = 84) tionality, 20 patients (62.5%) had non-functional tumors mean tumor size was 51mm (SD = 46.3) in contrast with and 12 (38.7%) functional tumors: 5 insulinomas (15.6%), 45.8mm (SD = 31.6) in the functional tumor group (n = 5 gastrinomas (15.6%), 1 glucagonoma (3.1%) and 1 GHo- 27) (p = 0.984). ma (3,1%). Five cases (15.6%) were located in the head,

Rev Esp Enferm Dig 2017;109(11):778-787 782 L. HURTADO-PARDO ET AL. Rev Esp Enferm Dig

Table 2. Comparison of non-functioning and functioning Table 3. Clinical and pathological characteristics cystic pancreatic neuroendocrine tumors of MEN-1 related cPNET Variable Non-functioning Functioning p Variable n (%) Gender, n (%) 0,99 Gender – Male 135 (51.5) 18 (51.4) – Male 14 (51.9) – Female 49.5 (15.8) 17 (48.6) – Female 13 (48.1) Age* mean (SD) 49.5 (15.8) 50.3 (16.4) 0.896 Age*, mean (SD) 41.7 (15) Genetic disorder, n (%) 0.001 Endocrine manifestations – Sporadic 240 (91.6) 23 (65.7) – Non-functioning 20 (62.5) – MEN-1 20 (7.7) 12 (34.3) – Insulinoma 5 (15.6) – Von Hippel Lindau 1 (0.4) 0 (0) – Gastrinoma 5 (15.6) – Tuberous sclerosis 1 (0.4) 0 (0) – Glucagonoma 2 (3.1) Location, n (%) 0.016 – GHRHoma 2 (3.1) – Head 75 (28.8) 8 (23.5) Location – Body 48 (18.5) 3 (8.8) – Head 5 (15.6) – Tail 130 (50) 17 (50) – Body 6 (18.8) – Multifocal 7 (2.6) 6 (17.6) – Tail 12 (37.5) Mean diameter (mm)† – Multifocal 9 (28.1) 51 (46.3) 45.8 (31.6) 0.983 (SD) Diameter (mm)† mean (SD) 40.23 (29.9) Ki-67, n (%) 1 Ki-67 – Low 134 (86.5) 15 (88.2) – Low 18 (100) – Intermediate 21 (13.5) 2 (11.7) Stage – High 0 (0) 0 (0) – I 7 (26.9) Regional lymph node 0.003 – IIa-IIb 16 (61.5) N, n (%) – IIIa-IIIb 2 (7.7) – N0 157 (90.8) 20 (69) – IV 1 (3.8) – N1 16 (9.2) 9 (31) Follow-up (months)‡ mean (SD) 46.8 (39.6) Stage, n (%) 0.007 Recurrence§ 7 (28) – I 41 (26.3) 5 (15.2) Death║ 1 (4) – IIa-IIb 95 (60.9) 17 (51.5) *Age: n = 27. †Diameter (mm): n = 26. ‡Follow-up in months: n = 25. §Recur- – IIIa-IIIb 14 (9) 4 (12.1) rence: n = 25. ║Death: n = 25 – IV 6 (3.8) 7 (21.2) *Age: functioning (n = 89), non-functioning (n = 28). †Diameter: functioning (n = 84), non-functioning (n = 27). DISCUSSION

cPNETs account for 13-17%% of pancreatic neuro- 6 (18.8%) in the body, 12 (37.5%) in the tail and 9 (28.1%) endocrine tumors and 9-10% of resected cystic tumors were multifocal. The mean diameter reported for 26 of the (1,2). Given their low incidence, their biological behavior patients was 40.2 mm (SD = 29.9). is largely unknown. Although the first reports date from According to the ENETS classification (n=26), 7 cases 1940, such tumors were not recognized as a specific entity (26.9%) were stage I, 16 (61.5%) stage II, 2 (7.7%) stage until 2008 (1,88). Most cases have been reported as isolat- III and 1 (3.8) stage IV. The Ki-67 index score was low ed cases, and only 13 authors have published series with in the 18 cases for whom this information was available. more than 10 cases (1,2, 77-87). Follow-up was recorded for 25 cases, with a mean of In the literature two well defined periods can be iden- 46.8 months (SD = 39.6) and a range of 1 to 120 and 7 tified: up until 2000 reports were of large symptomatic relapses (28%) and 1 death (4%) occurred. tumors and from 2000 onwards there was a predomi- Overall survival at 1, 5 and 10 years in this group of patients nance of small tumors, diagnosed incidentally due to the with MEN-1 was 100%, 91.7% and 91.7%, respectively. DFS use of more sensitive imaging techniques, as a result of at 1, 5 and 10 years was 100%, 60% and 60%, respectively. which there has been an increase in pancreatic “inciden-

Rev Esp Enferm Dig 2017;109(11):778-787 2017, Vol. 109, N.º 11 CYSTIC PANCREATIC NEUROENDOCRINE TUMORS (cPNETs): A SYSTEMATIC REVIEW 783 AND META-ANALYSIS OF CASE SERIES talomas” and their management has generated certain In our review, we found no differences between sex- controversy (89). es. The mean age of presentation was 53.2 years (SD = On the occasion of treating 5 cases in our center, and 16.1) but was significantly higher in men (57.1 vs. 49.5; with the aim of investigating the phenotype and oncolog- p = 0.001). Of note, 135 cases (44.6%) were diagnosed ic outcomes of these tumors, we performed a systematic incidentally, a figure consistent with that from the largest review and a meta-analysis of published cases. series of solid tumors (3,5). Abdominal pain (35%) was The etiopathogenesis of cPNETs is controversial. Kami- the most frequent symptom. Other phenotypical features of sawa et al. (62) proposed that the slow, expansive growth the tumors described are that 87.1% are sporadic and only of neuroendocrine tumors led to the development of a 7.74% are hereditary and that 85% were non-functional fibrous capsule compromising blood flow to the tumors tumors, figures that are higher than those reported for solid and giving rise to infarction and central necrosis. Similarly, tumors (1,2). Another striking characteristic is that the size Buetow et al. (87) in a series of 133 pancreatic neuroendo- of non-functional and functional tumors was similar (51 crine tumors concluded that the presence of cystic degener- mm vs. 45.8 mm; p = 0.983) when it is well known that in ation or necrosis was correlated with the size of the tumor. solid tumors non-functional tumors are generally larger. In contrast, Iacono et al. (53) suggested that hemorrhage is This could be due to the more rapid growth of such tumors the initial event in the development of the tumor’s cystic because of their hemorrhagic or necrotic component of form. Other authors propose that cPNETs are caused by the cPNETs (62,91). intraductal growth of a neuroendocrine tumor. According to the ENETS classification, most tumors Furthermore, pathologic staging of pancreatic neuro- correspond to initial stages limited to the pancreas (I, IIa endocrine tumors was not published until 2010 (ENETS) and IIb), with a good prognosis following resection and a (8), which also led to confusion until the classification was 1, 5 and 10 year DFS of 98.5%, 91.5% y 91.5%, respec- widely adopted. tively. In more advanced stages (IIIb and IV) 1, 5 and 10 X-ray diagnosis has low specificity as the images are year DFS was significantly lower: 100%, 54.2% and 54.2% similar to those of other cystic lesions of the pancreas (p = 0.001) (Fig. 2). The ki-67 index was low in 85.3% of such as solid pseudopapillary tumors, mucinous tumors, cases and intermediate in 14.9%, figures similar to those intraductal mucinous papillary neoplasms and pancreatic reported for solid tumors. These data confirm the prog- metastases, among others. In the review by Singhi et al. nostic value of the ENETS and WHO classifications for (1) 43% of cPNETs were wrongly diagnosed. Khashab et cPNETs (92). al. (90) reported the diagnostic superiority of ultrasound When we analyzed the series according to the degree endoscopy with an increase in yield over computed tomog- of functionality, we observed no differences in sex and raphy of 36% and over nuclear magnetic resonance imag- mean age of presentation. A larger proportion of patients ing of 54%. with MEN-1 was observed in the functional tumor group In their review, Morales-Oyarvide et al. (81) reported (34.3% vs. 7.3%; p=0.001) and a significantly greater pro- a sensitivity of 71% using cytology as opposed to a sensi- portion of multifocal tumors (17.6% vs. 2.6%; p=0.016) in tivity of 38% when only using endoscopic ultrasound and the functional tumor group as compared to the non-func- concluded that cytologic diagnosis with FNPA is the most tional group. When ENETS stage was assessed, we found appropriate diagnostic test. In our review, we obtained sim- a higher percentage of functional tumors in stage IV as ilar data, with a diagnostic sensitivity when using FNPA of compared to non-functional tumors (21.2% vs. 3.8%; p = 78.5%, a figure similar to that reported by Singhi, Chen, 0.007). DFS was significantly lower in functional tumors Morales-Oyarvide and Muthusamy (1,79,81,90). (61.9% at 6 years) as compared with non-functional tumors Although the management of cPNETs has generally (91.7% at 5 years) (p = 0.001); which is not consistent with been surgical, the increased sensitivity and greater use of the experience reported for solid neuroendocrine tumors imaging techniques has led to a rise in the incidental diag- (3,93). Perhaps these differences may be explained by the nosis of cPNETs, and as a result their treatment has become significantly greater proportion of multifocal and MEN-1 a matter of controversy especially in elderly patients with tumors in the functional group (p = 0.016 and p = 0.001 a high surgical risk. The therapeutic decision will depend respectively). on factors such as potential malignancy, life expectancy, cPNETs may present as sporadic tumors or in the context tumor location, the presence of symptoms and the expe- of hereditary tumors. The analysis of hereditary cPNETs rience of the surgeon. Ligneau et al. (86) proposed that focused on multiple endocrine neoplasms type 1 (MEN-1) cPNETs should be treated with surgical resection given as reports of other entities (Von-Hippel Lindau syndrome their possible malignant progression. or Neurofibromatosis type 1) are anecdotal. Ligneau et al. Cloyd et al. (77) were the first to classify PNETs depend- (86) report that the phenotype of cPNETs associated with ing on their cystic component and concluded that purely MEN-1 syndrome is clinically and pathologically differ- cystic PNETs represent a subgroup that may be monitored ent to that of sporadic cPNETs as they present in younger clinically and radiologically without immediate resection, patients, with a greater percentage of functional tumors given their benign progression. which are multiple and generally located in the tail of the

Rev Esp Enferm Dig 2017;109(11):778-787 784 L. HURTADO-PARDO ET AL. Rev Esp Enferm Dig

pancreas. In our review there were no differences with In spite of the biologic heterogeneity of cPNETs, the regard to sex; the mean age of presentation was 41.7 years pathologic classification as established by ENETS effec- –lower than in sporadic PNETs due to the high frequency tively predicts long-term prognosis in these patients, which of functional tumors and possible to the better and longer confirms the prognostic value of this classification. follow-up in patients with this syndrome. The incidence of non-functional cPNET tumors in this subgroup (62.5%) is similar to that reported in other series of patients with REFERENCES MEN-1, although great variability exists in the literature (5,77,94,95). The mean size (42 mm) is not very represen- 1. Singhi AD, Chu LC, Tatsas AD, et al. Cystic Pancreatic Neuroendo- crine Tumors. Am J Surg Pathol 2012;36(11):1666-73. DOI: 10.1097/ tative as they are surgical series they reflect the clinical PAS.0b013e31826a0048 guidelines to treat tumors ≥ 2 cm (5,94). 2. Gaujoux S, Tang L, Klimstra D, et al. The outcome of resected cystic Disease-free survival in hereditary tumors was lower pancreatic endocrine neoplasms: A case-matched analysis. Surgery than that reported for sporadic tumors due to the presence 2012;151(4):518-25. DOI: 10.1016/j.surg.2011.09.037 3. Ellison TA, Wolfgang CL, Shi C, et al. A single institution’s 26-year of multifocal tumors and lymph node metastases, a finding experience with nonfunctional pancreatic neuroendocrine tumors: a in line with that reported by other authors (5,77,95). validation of current staging systems and a new prognostic nomogram. We are aware of the limitations of this study such as Ann Surg 2014;259(2):204-12. DOI: 10.1097/SLA.0b013e31828f3174 4. Birnbaum DJ, Gaujoux S, Cherif R, et al. Sporadic nonfunction- the great heterogeneity of the data reported in the case ing pancreatic neuroendocrine tumors: prognostic significance of reports and in the case reviews and inter-study variability. incidental diagnosis. Surgery 2014;155(1):13-21. DOI: 10.1016/j. However, the information reported in case report is more surg.2013.08.007 accurate than in large case reviews. 5. Crippa S, Partelli S, Bassi C, et al. Long-term outcomes and prognostic factors in neuroendocrine carcinomas of the pancreas: Mor- Secondly, this review only includes cases of cPNETs phology matters. Surgery 2016;159(3):862-71. DOI: 10.1016/j. that were resected and histologically confirmed, and thus surg.2015.09.012 the real prevalence of cPNETs may not be fully reflected. 6. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for Furthermore, the cystic morphology seems to be a marker systematic reviews and meta-analyses: The PRISMA statement. Int J Surg 2010;8(5):336-41. DOI: 10.1016/j.ijsu.2010.02.007 of lower biologic aggressiveness, reducing the number of 7. Gagnier JJ, Kienle G, Altman DG, et al. The CARE guidelines: con- surgical resections in this type of tumor. sensus-based clinical case report guideline development. J Diet Suppl Thirdly, given the size of the sample and the lack of 2013;10(4):381-90. DOI: 10.3109/19390211.2013.830679 8. Klöppel G, Rindi G, Perren A, et al. The ENETS and AJCC/UICC TNM event or loss of follow-up in the patients, it was decided classifications of the neuroendocrine tumors of the gastrointestinal tract not to undertake a multivariate analysis to determine inde- and the pancreas: a statement. Virchows Arch 2010;456(6):595-7. DOI: pendent associations. 10.1007/s00428-010-0924-6 In contrast to these limitations, case reports and 9. Ricci C, Casadei R, Taffurelli G, et al. Validation of the 2010 WHO classification and a new prognostic proposal: A single centre retrospec- case series are a valuable source of information for evi- tive study of well-differentiated pancreatic neuroendocrine tumours. dence-based medicine investigating disorders which are Pancreatol Off J 2016;16(3):403-10. DOI: 10.1016/j.pan.2016.02.002 not very prevalent as is our case and which are not subsid- 10. Pergolini I, Iliopoulos O, Ferrone CR, et al. Clinico-pathological iary to randomized studies (6,7,96-99). features and survival of resected pancreatic neuroendocrine tumors associated with von Hippel Lindau disease and multiple endocrine neo- To our knowledge, the current review represents the plasia type 1. Gastroenterology 2016;150(4):S1232-3. DOI: 10.1016/ largest number of patients with cPNETs reported in the S0016-5085(16)34168-3 English literature. 11. Amagai H, Jingu K, Kitabayashi H, et al. A case of nonfunctioning neuroendocrine tumor of the pancreas which had difficulty in differ- entiation with mucinous cystic tumor. Chiba Med J 2016;92(2):45-50. 12. Kawai G, Shirota G, Unno T WT. A case of the NET of the pancreas CONCLUSIONS that was difficult to distinguish from SPT. Japanese J Clin Radiol 2016;61(6):777-81. 13. Tsujino T, Samarasena J CK. Endoscopic ultrasound-guided, nee- Cystic pancreatic neuroendocrine tumors are an uncom- dle-based confocal laser endomicroscopy in the diagnosis of cystic mon entity which in 44.5% of cases present incidentally. pancreatic neuroendocrine tumor: Emerging imaging criteria from They represent an entity that is both clinically and patho- surgical pathologic correlation in two cases. Am J Gastroenterol 2015;110(1):S107-8. logically different to solid tumors and must be included in 14. Lico V, Milanetto AC, Moletta L, et al. Pancreatic neuroendocrine tumor the differential diagnosis of cystic lesions of the pancreas. and ileal carcinoid i nacromegaly. Pluriglandular association in non- It seems advisable to follow the same criteria in staging MEN1 patient: a case report. Neuroendocrinology 2015;102(1-2):151. and resectability (≤ 2 cm) in both sporadic and hereditary 15. Humphrey PE, Alessandrino F, Bellizzi AM, et al. Non-hyperfunction- ing pancreatic endocrine tumors: multimodality imaging features with (MEN-1) tumors. Endoscopic ultrasound in association histopathological correlation. Abdom Imaging 2015;40(7):2398-410. with FNPA has a high sensitivity and specificity. DOI: 10.1007/s00261-015-0458-0 Unlike solid tumors, functional cPNETs have a lower 16. Apodaca-Torrez FR, Oliveira ML, Triviño T, et al. Tumor neuroen- overall survival possibly due to the greater incidence of docrino quístico no funcionante del páncreas. Una presentación poco usual. Cir Esp 2015;93(1):43-54. MEN-1 in functional cystic neuroendocrine tumors. In 17. Welch A, McGarrity T, Mani H MM. Cystic neuroendocrine tumor of functional cPNETs, the diagnosis of MEN-1 should be the pancreas: a rare type of pancreatic cystic lesion accurately diag- ruled out. nosed and staged by EUS-FNA. Pract Gastroenterol 2013;37(10):31-4.

Rev Esp Enferm Dig 2017;109(11):778-787 2017, Vol. 109, N.º 11 CYSTIC PANCREATIC NEUROENDOCRINE TUMORS (cPNETs): A SYSTEMATIC REVIEW 785 AND META-ANALYSIS OF CASE SERIES

18. Bombardieri R, Moavero R, Roberto D, et al. Pancreatic neu- 41. Christein JD MK. Pancreatic somatostitoma associated with roendocrine tumor in a child with a tuberous sclerosis complex 2 obstructive jaundice, pancreatic and biliary Kaposi´s sarcoma (TSC2) mutation. Endocr Pract 2013;19(5):e124-8. DOI: 10.4158/ in a man with neurofibromatosis. HPB 2003;5(1):49-53. DOI: EP13010.CR 10.1080/13651820310003575 19. Addeo A, Bini R, Viora T, et al. Von Hippel-Lindau and myotonic 42. Vandecaveye V, Verswijvel G, Colla P, et al. Cystic insulinoma of the dystrophy of Steinert along with pancreatic neuroendocrine tumor pancreas in a patient with myotonic dystrophy: correlation of imaging and renal clear cell carcinomal neoplasm: Case report and review of and pathologic findings. JBR-BTR 2003;86(5):268-71. the literature. Int J Surg Case Rep 2013;4(8):648-50. DOI: 10.1016/j. 43. Ahrendt SA, Komorowski RA, Demeure MJ, et al. Cystic pancreatic ijscr.2013.03.004 neuroendocrine tumors: is preoperative diagnosis possible? J Gastro- 20. Deshmukh SD, Gulati HK, Gaopande V, et al. Incidental cystic endo- intest Surg 2002;6(1):66-74. crine tumor of the pancreas: a case report with immunohistochemi- 44. Oberg KC, Wells K, Seraj IM, et al. ACTH-secreting islet cell tumor cal study. J Cancer Res Ther 2012;8(2):289-91. DOI: 10.4103/0973- of the pancreas presenting as bilateral ovarian tumors and Cush- 1482.98992 ing’s syndrome. Int J Gynecol Pathol 2002;21(3):276-80. DOI: 21. Hu K-C, Chang W-H, Chu C-H, et al. Pancreatic head well-differenti- 10.1097/00004347-200207000-00012 ated endocrine tumor. Am J Surg 2011;201(3):e26-8. DOI: 10.1016/j. 45. Kato K, Kondo S, Ambo Y, et al. Nonfunctioning endocrine tumor of amjsurg.2010.03.019 the pancreas with extrapancreatic growth and cyst formation: report of 22. Potru R, Collins C S V. Rare presentation of cystic nonfunctioning a case. Surg Today 2000;30(7):651-4. DOI: 10.1007/s005950070107 pancreatic neuroendocrine tumor: a case-series. Am J Gastroenterol 46. Brown K, Kristopaitis T, Yong S, et al. Cystic glucagonoma: A rare 2011;106(2):S219-20. variant of an uncommon neuroendocrine pancreas tumor. J Gastrointest 23. Baig M, Chin U, Wilhite D CH. Cystic pancreatic neuroendocrine Surg 2(6):533-6. tumors. Am J Gastroenterol 2011;106(2):S213. 47. Kotoulas C, Panayiotides J, Antiochos C, et al. Huge non-functioning 24. Higashi T, Kawaguchi Y, Tsukune Y, et al. A case report of glucagon- pancreatic cystic neuroendocrine tumour: a case report. Eur J Surg producing multiple pancreatic neuroendocrine tumors that involved Oncol 1998;24(1):74-6. DOI: 10.1016/S0748-7983(98)80133-9 pancreatic cystic lesions. Pancreatology 2010;10(1):70. 48. Sohaib SA, Reznek RH, Healy JC, et al. Cystic islet cell tumors of 25. Diehl DL, Blansfield J, Li J. Cystic pancreatic neuroendocrine tumor. the pancreas. AJR Am J Roentgenol 1998;170(1):217. DOI: 10.2214/ Gastrointest Endosc 2010;71(6):1064-5; discussion 1065. DOI: ajr.170.1.9423637 10.1016/j.gie.2009.12.016 49. Sarui H, Yoshimoto K, Okumura S, et al. Cystic glucagonoma with loss 26. Colović RB, Matić S V, Micev MT, et al. Two synchronous somatostati- of heterozygosity on chromosome 11 in multiple endocrine neoplasia nomas of the duodenum and pancreatic head in one patient. World J type 1. Clin Endocrinol (Oxf) 1997;46(4):511-6. DOI: 10.1046/j.1365- Gastroenterol 2009;15(46):5859-63. DOI: 10.3748/wjg.15.5859 2265.1997.1380965.x 27. Kobayashi Y, Yamao K, Sawaki A, et al. A case of pancreatic neu- 50. Zanow J, Gellert K, Benhidjeb T MJ. Cystische tumoren des pankreas. roendocrine tumor (NET) with atypical looking which diagnosed by Chirurg 1996;67:719-24. EUS-FNA. Pancreas 2009;38(8):1018. 51. Goto M, Nakano I, Sumi K, et al. Cystic insulinoma and nonfunction- 28. Basu A, Sistla SC IK. Image of the month. Arch Surg 2009;144(1):89. ing islet cell tumor in multiple endocrine neoplasia type 1. Pancreas DOI: 10.1001/archsurg.2008.527-a 1994;9(3):393-5. DOI: 10.1097/00006676-199405000-00020 29. Kamei K, Yasuda T, Shinzaki W, et al. Cystic nonfunctioning pancreat- 52. Weissmann D, Lewandrowski K, Godine J, et al. Pancreatic cystic ic endocrine neoplasm presenting communication with main pancreatic islet-cell tumors. Clinical and pathologic features in two cases with duct. Dig Surg 2009;26(1):25-6. DOI: 10.1159/000193328 cyst fluid analysis. Int J Pancreatol 1994;15(1):75-9. 30. Charfi S, Marcy M, Bories E, et al. Cystic pancreatic endocrine tumors: 53. Iacono C, Serio G, Fugazzola C, et al. Cystic islet cell tumors of the an endoscopic ultrasound-guided fine-needle aspiration biopsy study pancreas. A clinico-pathological report of two nonfunctioning cases with histologic correlation. Cancer 2009;117(3):203-10. DOI: 10.1002/ and review of the literature. Int J Pancreatol 1992;11(3):199-208. DOI: cncy.20024 10.1007/BF02924187 31. Chetty R, El-Shinnawy I. Intraductal pancreatic neuroendocrine tumor. 54. Yagihashi S, Yagihashi N, Nagai K. Cystic Pancreatic Glucagonoma Endocr Pathol 2009;20(4):262-6. DOI: 10.1007/s12022-009-9093-z in Contact with Insulinoma Found in a Hypoglycemic Patient. Pathol 32. Kongkam P, Al-Haddad M, Attasaranya S, et al. EUS and clinical - Res Pract 1992;188(6):751-6. DOI: 10.1016/S0344-0338(11)80173-1 characteristics of cystic pancreatic neuroendocrine tumors. Endoscopy 55. Nojima T, Kojima T, Kato H, et al. Cystic endocrine tumor of the 2008;40(7):602-5. DOI: 10.1055/s-2007-995740 pancreas. Int J Pancreatol 1991;10(1):65-72. 33. Ballarin R, Masetti M, Losi L, et al. Cystic pancreatic neuroendocrine 56. Azoulay D, Sauvanet A, Bonnichon P, et al. Non-functioning tumors neoplasms with uncertain malignant potential: report of two cases. of the endocrine pancreas. Concept, diagnosis, treatment. J Chir (Paris) Surg Today 2009;39(2):162-7. DOI: 10.1007/s00595-008-3806-7 1990;127(4):185-90. 34. Sumarac-Dumanovic M, Micic D, Krstic M, et al. Pitfalls in diag- 57. Warshaw AL, Compton CC, Lewandrowski K, et al. Cystic tumors of nosing a small cystic insulinoma: a case report. J Med Case Rep the pancreas. New clinical, radiologic, and pathologic observations in 2007;1(1):181. DOI: 10.1186/1752-1947-1-181 67 patients. Ann Surg 1990;212(4):432-43-5. DOI: 10.1097/00000658- 35. Kajiwara M, Gotohda N, Konishi M, et al. Cystic endocrine tumor of 199010000-00006 the pancreas with an atypical multilocular appearance. J Hepatobiliary 58. von Herbay A, Sieg B, Otto HF. Solid-cystic tumour of the pancreas. Pancreat Surg 2007;14(6):586-9. DOI: 10.1007/s00534-006-1218-x An endocrine neoplasm? Virchows Arch A Pathol Anat Histopathol 36. Li Destri G, Reggio E, Veroux M, et al. A rare cystic non-functioning 1990;416(6):535-8. neuroendocrine pancreatic tumor with an unusual presentation. Tumori 59. Davtyan H, Nieberg R, Reber HA. Pancreatic cystic endocrine 92(3):260-3. neoplasms. Pancreas 1990;5(2):230-3. DOI: 10.1097/00006676- 37. Goh BKP, Ooi LLPJ, Tan YM, et al. Clinico-pathological features of 199003000-00017 cystic pancreatic endocrine neoplasms and a comparison with their sol- 60. Lapeyrie H, Loizon P, Chapuis H, et al. Non-functional or silent endo- id counterparts. Eur J Surg Oncol 2006;32(5):553-6. DOI: 10.1016/j. crine tumors of the pancreas. Apropos of a case of cystic form. Ann ejso.2006.02.017 Chir 1989;43(4):302-5. 38. Tamagno G, Maffei P, Pasquali C, et al. Clinical and diagnostic aspects 61. Yagihashi S, Sato I, Kaimori M, et al. Papillary and cystic tumor of of cystic insulinoma. Scand J Gastroenterol 2005;40(12):1497-501. the pancreas. Two cases indistinguishable from islet cell tumor. Cancer DOI: 10.1080/00365520510024160 1988;61(6):1241-7. DOI: 10.1002/1097-0142(19880315)61:6<1241::AID- 39. Imaoka H, Yamao K, Salem AAS, et al. Pancreatic endocrine neo- CNCR2820610631>3.0.CO;2-T plasm can mimic serous cystadenoma. Int J Gastrointest Cancer 62. Kamisawa T, Fukayama M, Koike M, et al. A case of malignant cystic 2005;35(3):217-20. DOI: 10.1385/IJGC:35:3:217 endocrine tumor of the pancreas. Am J Gastroenterol 1987;82(1):86-9. 40. Gerke H, Byrne MF, Xie HB, et al. A wolf in sheep’s clothing: a non- 63. Stair JM, Schaefer RF, McCowan TC, et al. Cystic islet cell tumor functioning islet cell tumor of the pancreas masquerading as a micro- of the pancreas. J Surg Oncol 1986;32(1):46-9. DOI: 10.1002/ cystic (serous cystic) adenoma. JOP 2004;5(4):225-30. jso.2930320113

Rev Esp Enferm Dig 2017;109(11):778-787 786 L. HURTADO-PARDO ET AL. Rev Esp Enferm Dig

64. Loiterman DA, Schwartz IS, Aufses AH. Cystic islet cell tumor: diag- a Less Aggressive Behavior. Neuroendocrinology 2010;92(4):246-51. nosis and surgical management. Mt Sinai J Med 1985;52(3):212-5. DOI: 10.1159/000318771 65. Heidrich W, Kubale R, Seitz G, et al. Endocrine tumor of the pancreas- 85. Baker MS, Knuth JL, DeWitt J, et al. Pancreatic cystic neuroendocrine -a case report on the differential diagnosis of cystic epigastric tumors. tumors: preoperative diagnosis with endoscopic ultrasound and fine- Ultraschall Med 1986;7(3):138-40. DOI: 10.1055/s-2007-1011932 needle immunocytology. J Gastrointest Surg 2008;12(3):450-6. DOI: 66. Pogany AC, Kerlan RK, Karam JH, et al. Cystic insulinoma. AJR Am 10.1007/s11605-007-0219-7 J Roentgenol 1984;142(5):951-2. DOI: 10.2214/ajr.142.5.951 86. Ligneau B, Lombard-Bohas C, Partensky C, et al. Cystic endocrine 67. Ho PW, Moore GW, Hoge AF. Glucagonoma occurring as a tumors of the pancreas: clinical, radiologic, and histopathologic large cystic abdominal mass. South Med J 1984;77(5):666. DOI: features in 13 cases. Am J Surg Pathol 2001;25(6):752-60. DOI: 10.1097/00007611-198405000-00036 10.1097/00000478-200106000-00006 68. Thompson NW, Eckhauser FE, Vinik AI, et al. Cystic neuroendocrine 87. Buetow PC, Parrino T V, Buck JL, et al. Islet cell tumors of the pan- neoplasms of the pancreas and liver. Ann Surg 1984;199(2):158-64. creas: pathologic-imaging correlation among size, necrosis and cysts, DOI: 10.1097/00000658-198402000-00004 calcification, malignant behavior, and functional status. AJR Am J 69. Grosdidier J, Duprez J, Delfosse J, et al. Nonfunctional benign Langer- Roentgenol 1995;165(5):1175-9. DOI: 10.2214/ajr.165.5.7572498 hans tumor of the cystic type. Sem Hop 1983;59(23):1773-6. 88. Bordeianou L, Vagefi PA, Sahani D, et al. Cystic Pancreatic Endocrine 70. Frija J, Schmit P, Vadrot D, et al. Non-secreting islet cell adenoma Neoplasms: A Distinct Tumor Type? J Am Coll Surg 2008;206(6):1154- of the pancreas evaluated by computed tomography and sonography. 8. DOI: 10.1016/j.jamcollsurg.2007.12.040 Report of a case. Eur J Radiol 1982;2(2):160-1. 89. Yang M, Ke NW, Zeng L, et al. Survival Analyses for Patients With 71. Caplan RH, Koob L, Abellera RM, et al. Cure of acromegaly by Surgically Resected Pancreatic Neuroendocrine Tumors by World operative removal of an islet cell tumor of the pancreas. Am J Med Health Organization 2010 Grading Classifications and American Joint 1978;64(5):874-82. DOI: 10.1016/0002-9343(78)90531-4 Committee on Cancer 2010 Staging Systems. Medicine (Baltimore) 72. Riddle MC, Golper TA, Fletcher WS, et al. Glucagonoma syndrome in 2015;94(48):e2156. DOI: 10.1097/MD.0000000000002156 a 19-year-old woman. West J Med 1978;129(1):68-72. 90. Muthusamy VR, Chandrasekhara V, Acosta RD, et al. The role 73. Deltz VE PM. Zystisches Insulom. Zbl Chir 10975;100:1521-4. of endoscopy in the diagnosis and treatment of cystic pancreatic 74. Winston JH. Malignant islet cell adenoma in a pancreatic cyst. Report neoplasms. Gastrointest Endosc 2016;84(1):1-9. DOI: 10.1016/j. of a case. J Natl Med Assoc 1965;57(3):203-4. gie.2016.04.014 75. Leger L, Canivet J, Mizon JP, et al. Cyst-like tumors of the islands 91. Iacono C, Verlato G, Ruzzenente A, et al. Systematic review of central of Langerhans. Apropos of 2 cases. J Chir (Paris) 1967;94(1):7-20. pancreatectomy and meta-analysis of central versus distal pancreatec- 76. Joshi RA, Probstein JG. Benign cystic islet-cell tumor of the head of tomy. Br J Surg 2013;100(7):873-85. DOI: 10.1002/bjs.9136 the pancreas; report of a case probably of delta-cell origin. AMA Arch 92. Koh YX, Chok AY, Zheng HL, et al. A systematic review and meta- Surg 1955;71(1):74-7. DOI: 10.1001/archsurg.1955.01270130076012 analysis of the clinicopathologic characteristics of cystic versus solid 77. Cloyd JM, Kopecky KE, Norton JA, et al. Neuroendocrine tumors of pancreatic neuroendocrine neoplasms. Surgery 2014;156(1):83-96.e2. the pancreas: Degree of cystic component predicts prognosis. Surgery DOI: 10.1016/j.surg.2014.03.026 2016;160(3):708-13. DOI: 10.1016/j.surg.2016.04.005 93. Cienfuegos JA, Rotellar F, Salguero J, et al. A single institutions 78. Mitra V, Nayar MK, Leeds JS, et al. Diagnostic performance of endo- 21-year experience with surgically resected pancreatic neuroendo- scopic ultrasound (EUS)/endoscopic ultrasound--fine needle aspiration crine tumors: an analysis of survival and prognostic factors. Rev (EUS-FNA) cytology in solid and cystic pancreatic neuroendocrine Española Enfermedades Dig 2016;108(11):689-96. DOI: 10.17235/ tumours. J Gastrointestin Liver Dis 2015;24(1):69-75. reed.2016.4323/2016 79. Chen L, Nassar A, Kommineni VT, et al. Endoscopic ultrasonography- 94. Thakker R V, Newey PJ, Walls G V, et al. Clinical practice guidelines guided fine-needle aspiration cytology of surgically confirmed cystic for multiple endocrine neoplasia type 1 (MEN1). J Clin Endocrinol pancreatic neuroendocrine tumors: a Mayo Clinic experience. J Am Metab 2012;97(9):2990-3011. DOI: 10.1210/jc.2012-1230 Soc Cytopathol 2015;4(6):335-43. DOI: 10.1016/j.jasc.2015.04.001 95. Kouvaraki MA, Shapiro SE, Cote GJ, et al. Management of pancreatic 80. Komminemi VT, Chen L, Zhang J, et al. Cystic pancreatic neuroendo- endocrine tumors in multiple endocrine neoplasia type 1. World J Surg crine tumors: Clinical and endoscopic ultrasound characteristics and 2006;30(5):643-53. DOI: 10.1007/s00268-006-0360-y accuracy of EUS-FNA. Am J Gastroenterol 2014;109(2):S92. 96. Vandenbroucke JP. In defense of case reports. Ann Intern Med 81. Morales-Oyarvide V, Yoon WJ, Ingkakul T, et al. Cystic pancreatic 2001;134(2 Pt 2):330-4. DOI: 10.7326/0003-4819-134-4-200102200- neuroendocrine tumors: The value of cytology in preoperative diagno- 00017 sis. Cancer Cytopathol 2014;122(6):435-44. DOI: 10.1002/cncy.21403 97. Agha RA, Fowler AJ, Lee S-Y, et al. Systematic review of the meth- 82. Ho HC, Eloubeidi MA, Siddiqui UD, et al. Endosonographic and cyst odological and reporting quality of case series in surgery. Br J Surg fluid characteristics of cystic pancreatic neuroendocrine tumours: 2016;103(10):1253-8. DOI: 10.1002/bjs.10235 a multicentre case series. Dig Liver Dis 2013;45(9):750-3. DOI: 98. Barkin JA, Barkin JS. Pancreatic cysts. Controversies, advances, diag- 10.1016/j.dld.2013.02.021 noses, and therapies. Pancreas 2017;46(6).735-741 DOI: 10.1097/ 83. Konukiewitz B, Enosawa T, Klöppel G. Glucagon expression in cystic MPA.0000000000000831 pancreatic neuroendocrine neoplasms: an immunohistochemical analysis. 99. Moris M, Wallace MB. Intraductal papillary mucinous neoplasms Virchows Arch 2011;458(1):47-53. DOI: 10.1007/s00428-010-0985-6 and mucinous cystadenomas: current status and recommendations. 84. Boninsegna L, Partelli S, D’Innocenzio MM, et al. Pancreatic Cystic Rev Esp Enferm Dig 2017;109(5):358-67. DOI: 10.17235/reed.2017. Endocrine Tumors: A Different Morphological Entity Associated with 4630/2016

Rev Esp Enferm Dig 2017;109(11):778-787