^ Context.-Pancreatic neuroendocrine tumors (PancNETs) are rare and tend to get overshadowed by their more prevalent and aggressive ductal adenocarcinoma counterparts. The biological behavior of PancNETs is unpredictable, and thus management is controversial. However, the new
Objective.-To report on the utility and limitations of fine-needle aspiration in the preoperative evaluation and management of PancNETs.
Design.- This was a retrospective review of our institutional tumor database from 2002 to 2012. There were 25 cases of PancNETs that were localized and staged by medical imaging and diagnosed by fine-needle aspiration.
Results.-Fourteen patients underwent laparotomy, with some requiring only limited surgery; 4 had laparoscopic resections; 4 were serially observed without surgical intervention; and another 3 were inoperable. After a mean follow-up of 37 months, more than half of the patients had no evidence of disease, including most of those who underwent minimally invasive surgery.
Conclusions.-Fine-needle aspiration is a useful diagnostic adjunct to medical imaging in the preoperative evaluation and management of PancNETs. However, there are limitations with regard to grading PancNETs using this technique.
(Arch Pathol Lab Med. 2014;138:896-902; doi: 10.5858/ arpa.2013-0300-OA)
Cross-sectional imaging, endoscopic ultrasound (EUS), and fine-needle aspiration (FNA) are important in the preoperative evaluation of pancreatic masses.1-4 Although the majority of pancreatic tumors are ductal adenocarcinomas, other neoplasms such as neuroendocrine tumors, acinar cell carcinomas, solid pseudopapillary neoplasms, cystic neoplasms, and metastases also occur.5,6 Pancreatic neuroendocrine tumors (PancNETs) are a particularly interesting group of neoplasms because they can present clinically with neuroendocrine manifestations or remain silent until their presence is heralded by mass effect. They can be single and sporadic or can be a component of several syndromes.2-4 The biological behavior of PancNETs is unpredictable, especially in the absence of locoregional spread or metastases. Thus, the management is controversial. However, there is now a growing consensus that the new
MATERIALS AND METHODS
Approval for the study was obtained from our institutional Research Ethics Board. Surgical and cytopathology records of our hospital between 2002 and 2012 were searched for cases in which FNA of the pancreas was carried out. Of 547 FNAs identified, 25 (4.6%) were diagnostic for PancNETs. The slides and cell blocks including immunohistochemical stains for these cases were available. These 25 FNA cases are the focus of this study.
Among 471 pancreatic surgical pathology specimens, 24 resections met the criteria for PancNET. Among these 24 surgical pathology cases, 7 did not have a prior FNA that was diagnostic for PancNET. Of these 7, 3 had failed percutaneous FNA attempts because of patient discomfort (n ¼ 2) and hemorrhage (n ¼ 1). Two of the 7 cases (percutaneous) were not sufficiently diagnostic of PancNET because of an inadequate cell block (differential including solid pseudopapillary tumor and well-differentiated adenocarcinoma), and another 2 had no preoperative FNA. Thus, in total there were 17 surgical pathology resections with a prior FNA, and this served as a further cross-check on the FNA diagnosis.
Percutaneous Ultrasound FNA
Of the 25 PancNET FNA specimens, 12 were obtained using percutaneous ultrasound-guided FNA by a radiologist, which was a prior standard procedure in our institute. After obtaining consent, the patients initially were scanned to identify the lesion and the proper access site. Once a suitable aspiration site was selected, the overlying skin was marked and prepared using standard sterile technique, and a local anesthetic was administered. Fine-needle aspiration was then performed with a transabdominal or posterior approach using either a 20- or 22-gauge
Endoscopic Ultrasound FNA
In the remaining 13 cases, EUS was used to localize and aspirate the tumors. Endoscopic ultrasound-guided FNA was performed by a single experienced endosonographer. All patients provided informed consent before undergoing the procedure. Following conscious sedation (with intravenous fentanyl and Diazemuls,
Slide and Cell Block Preparation
A trained cytotechnologist went to the biopsy suite to provide rapid on-site adequacy evaluation of the specimen and to ensure the proper handling of material for cytologic diagnosis. During this process, a small drop of specimen was expelled onto one glass slide. The second glass was inverted and placed face down on the drop, and the slides were pulled apart. One slide was immediately placed in a Coplin jar containing 95% alcohol. The other slide was air dried, to be stained later with Diff-Quik stain in the cytology laboratory. The alcohol-fixed slide was stained with the Diff-Quik stain, a coverslip was placed, and the slide was examined under the microscope. The presence of 5 or 6 groups of cells deemed to represent the lesion was considered adequate. After completion of the adequacy assessment, the Diff-Quik-stained slide was put back in the Coplin jar containing 95% alcohol to decolorize, and then it was submitted for permanent Papanicolaou staining in the cytology laboratory. When adequate smears had been obtained, the remaining specimen was evaluated to determine if there was sufficient material for cell block preparation. In most circumstances, additional samples for the cell block were obtained to provide material for ancillary testing. This material was immediately fixed in 10% neutral buffered formalin.
Immunocytochemistry and Immunohistochemistry
Immunocytochemistry was carried out on the cell blocks for all 25 FNA cases. They were stained for the neuroendocrine markers synaptophysin, chromogranin A, and CD56. In select cases, additional stains for neurohormonal, epithelial, or lymphoid markers or vimentin were run as part of the diagnostic workup for the FNAs. For the 17 surgical resections, a representative block from each case was stained for the neuroendocrine markers. In addition, for both the cytology and resection specimens, a cell block or representative tissue block was stained for Ki67 (MM1, (Vision BioSystems,
Patient Clinical Characteristics
Of the 25 patients, there were 13 men and 12 women (Table 1). The age range was 37 to 82 years, with a mean of 61 years. A percutaneous FNA was performed on 12 patients, and EUS on the other 13. Medical imaging studies showed that the most common location of the tumor was the pancreatic head (n ¼ 11; Figure 1) with lesser numbers in the tail (n ¼ 5; Figure 2), body (n ¼ 3), uncinate process (n ¼ 3), and neck (n ¼ 3). The tumor maximum dimension ranged from 1.0 to 7.2 cm, with a mean of 3.3 cm.
Cytomorphologic and Immunocytochemical Findings
The aspirates were generally quite cellular with a preponderance of single cells, but also contained loosely cohesive overlapping aggregates or flat groups of cells. The cells were relatively uniform, with a modest amount of pale cytoplasm and indistinct cell borders (Figures 3 and 4). Their nuclei were eccentric/plasmacytoid and mostly round or oval with a salt-and-pepper chromatin pattern. Nucleoli were small or inconspicuous. Mitotic figures were rarely encountered. The background contained a variable amount of blood and normal cells such as gastric, duodenal, or pancreatic epithelial, acinar, or ductal cells, depending on the manner in which the FNA had been carried out. In all cases, the neuroendocrine nature of the tumor cells on the cell block was confirmed using immunocytochemistry by at least 2 of 3 pan-neuroendocrine markers (Figure 5).
The surgical resections showed the organoid pattern characteristic of neuroendocrine tumors, and most of them were well differentiated, as shown in Table 1: the Ki67 labeling index was WHO grade 1 (,3%) in 7 of 17 cases, and grade 2 (3%-20%) in 9 of 17 cases. This generally corresponded to the relatively uniform cytomorphology of the cases. We were unable to distinguish grade 1 from grade 2 tumors on the histology alone. There was, however, 1 case of 17 that turned out to be WHO grade 3 (.20%) on the resection specimen (case 23). Although it showed a trabecular and insular organoid pattern, there were also scattered foci of tumor necrosis, and the mitotic count was 12 per 10 high-power fields. The cytomorphology of this case showed more variability in nuclear size and shape, with some cells having a higher nuclear/cytoplasmic ratio, and had 9 mitotic figures out of a total of approximately 5300 cells that were counted on the cell block (Figure 6). No areas of necrosis were seen in any of the cases, but the latter case showed a few scattered foci of apoptotic debris. We also attempted to determine the Ki67 labeling index on the cell blocks, but we encountered a number of challenges, as follows: (1) It was sometimes difficult to confidently distinguish PancNET cells from nonneoplastic contaminant cells such as acinar, epithelial, or lymphoplasmacytoid cells of pancreatic or gastrointestinal origin on cell block sections stained with Ki67; (2) In some instances, especially in the percutaneous FNAs, there were fewer than 500 cells on the sections (Table 1), 500 being the suggested minimum threshold.7-9 Furthermore, because the cell blocks were just a cell button of the needle rinse, there was considerable variation in the cells on the section as one cut deeper into the block (ie, section-to-section variation); (3) Another constraint was the total number of sections that could be cut from a cell block, ranging from 8 to more than 20. Indeed in one case, the cell block was exhausted of cells at the time of recutting for Ki-67; (4) We could not as easily determine hot spots on the cell block as we did on tissue sections, and therefore ended up counting all the cells (to the nearest 100) that we deemed to be tumor cells on the cell block (Table 1).
Despite the shortcomings in determining the Ki67 labeling index on FNA material, we went ahead and compared the grading of the resections with that of FNA material in the 16 cases on which this was feasible (Table 1). In 7 of 16 cases (44%), the cytologic Ki67 index corresponded to that of the resections. However, in 9 of 16 cases (56%) there was a discrepancy: 5 (31%) were upgraded on the resection from grade 1 to 2, and 4 (25%) were downgraded from grade 2 to 1 (Table 1).
Fourteen of 25 patients (56%) underwent laparotomy (Table 2). Six of them had a distal pancreatectomy and splenectomy, 4 required a pancreaticoduodenectomy, and in another 3 the tumors were enucleated. In 1 patient, the tumor was deemed to be unresectable at exploratory laparotomy and only core biopsies of the tumor were taken.
Four of 25 patients (16%) underwent laparoscopy, and in all 4 a distal pancreatectomy and splenectomy was carried out; in 3 of the 4, it was a hand-assisted laparoscopic procedure.
Three of 25 patients (12%) were deemed inoperable based on imaging evaluations demonstrating local invasion of major vessels.
Another 4 patients (16%) had surgically resectable tumors. However, 3 elderly patients with relatively small, nonfunctional tumors opted for a wait-and-see approach, and a fourth patient with significant comorbidities was considered medically inoperable and was monitored through serial observations.
From the total group of 25 patients, 6 patients with locoregional or liver metastases also received additional forms of treatment, mostly chemotherapy and/or radiotherapy.
The follow-up period ranged from 4 to 96 months, with a mean follow-up of 37 months (Table 1). More than half of the patients (14 of 25) were alive and well with no evidence of disease. A little more than a quarter of the patients (7 of 25) were alive with disease, including all 4 patients who were observed, 1 patient who was inoperable, and 2 other patients who were originally staged as T2N0M0 and T3N1M0 respectively but subsequently developed liver metastasis. Three patients (3 of 25) died of disease; all 3 were inoperable based on medical imaging findings (n ¼ 2) or at laparotomy (n ¼ 1). One patient (1 of 25) was lost to follow-up after 84 months. The overall survival of those patients who underwent resection and those who only had a biopsy are shown in Figure 7.
Because the vast majority of solid pancreatic tumors are invasive ductal adenocarcinoma, this has almost become synonymous with cancer of the pancreas. As a result, other pancreatic neoplasms with significantly different biological behaviors are often overlooked in the analysis of research databases.5,6 One such tumor group is PancNETs. In this study, we showed that FNA can be used to accurately diagnose PancNET in most cases. A review of the literature2,14-20 shows that FNA compares favorably with core biopsies because the latter are technically more difficult, especially with smaller lesions, and also more limited in their capacity to sample the tumor. Only 3 cases during the time frame of this study were complicated by patient discomfort or hemorrhage, and all of them were during the era of the percutaneous approach. We had no instances of pancreatitis, infection, perforation, or needletrack tumor seeding. This low incidence of complications is along the lines reported by others and partly explains why FNA, especially EUS-FNA, has become the gold standard relative to core biopsies.2,14-20
The establishment of a correct diagnosis using FNA, when combined with the localization and staging that can be achieved through medical imaging, provides important information in the preoperative evaluation and management of these patients. For example, nonfunctional PancNETs confined to pancreas in the elderly are appropriately considered for observational management with periodic cross-sectional imaging instead of surgical intervention. The 3 patients in our series who were managed this way had a mean age of 80 years. All of them are still alive with no evidence of metastases after a mean follow-up of 21 months, though in 1 case the tumor has increased in size from 2.9 to 4.3 cm during a period of 31 months.
Four of our patients underwent laparoscopic surgery, all achieving negative surgical resection margins. All these patients are alive and well with no evidence of disease after a mean follow-up of 42.8 months. Although laparoscopy has been routinely carried out in general surgery since the 1980s, its use in the more complex pancreatic surgery has been a more recent development. However, more opportunities for the laparoscopic approach are arising for indolent tumors.1,10-13 In some series of laparoscopic pancreatic resections, PancNET histology comprises the largest proportion.10,13
In 3 patients, an enucleation of the tumor was carried out, a procedure generally considered when the morphology is known to be a PancNET rather than the more aggressive ductal adenocarcinoma. Fine-needle aspiration allows the surgical team to carry out a simpler, parenchyma-preserving operation with all the attendant benefits for the patient.
Even with PancNET, there is still a role for pancreaticoduodenectomy, depending on the location and stage of the tumor. However, the proportion of cases in which this operation is carried out is significantly less thanks to the improvements in medical imaging, FNA, and minimally invasive surgery.
Although FNA is generally associated with a higher diagnostic yield and fewer complications compared to core biopsies, one of its drawbacks is that the tumor cells are mostly in the form of single cells or small dispersed groups of cells, which puts some limitations in assessing tumor grade. The WHO grading of PancNETs is based on number of mitotic counts per unit area as measured in square millimeters or per high-power field(s). Although mitoses, and even necrosis, can be appreciated in FNA material, accurate quantification is curtailed by the disruption of pristine tissue architecture in FNAs. Similarly, the determination of Ki67 index may be impacted by the disruption of tissue architecture, because this makes it more challenging to define hot spots.7-9 Furthermore, it may be difficult to distinguish neuroendocrine cells from contaminant cells, especially in the absence of tissue architecture,21,22 on cell block sections stained with Ki67 (Figures 8 and 9). Because these contaminant cells are mostly physiologically labile cells, a significant proportion of them tend to stain for Ki67. Most indolent tumors are generally characterized by a low Ki-67 labeling index compared to high-grade tumors. Therefore, the effect of even small numbers of the contaminant cells on PancNET cell block sections could significantly influence tumor grade.7-9 In addition, despite the fact that 2000 tumor cells are recommended for accurate tumor grading, the minimum threshold of 500 cells is sometimes not attained in cell block sections.7-9 However, going forward, with the increasing utilization of EUS-FNA, the latter problem can ameliorated by requesting additional passes at the time of rapid on-site adequacy assessment.
In this study, 9 of 17 cases (56%) had a discrepancy in grading between the surgical resection and the cytology cases (5 cases were upgraded on the resection from grade 1 to 2, and 4 were downgraded from grade 2 to 1). We looked at the relationship of Ki67 grade (cytology and resections) with survival and metastasis, but we had insufficient numbers for statistical analysis. However, given the abovedescribed cytology limitations, we are inclined to put more weight on the grading based on the resections. The discrepancy rate between the cytology and resections could have been improved by using 5% as the cutoff for separating grade 1 from grade 2 lesions, as has been recently suggested,23,24 but a preponderance of studies7-9 appear to validate the current WHO classification.
It would therefore appear that there are still questions regarding the grading of PancNETs based on FNA material. Indeed, this is reflected by the conflicting reports in the literature.7-9,16,20,24-27 The relative lack of confidence in grading PancNETs based on FNA material (vis-a` -vis tissue) can therefore be regarded as a limitation. For example, in the 4 patients who were managed by the observational approach in our series, it would have been more reassuring to also know the tumor grade because this has a bearing on outcome. Likewise, in the surgical intervention group, tumor grade may influence the type of operation that is carried out. Furthermore, if systemic therapy is being contemplated in the setting of locally advanced or metastatic disease, the choice of therapy differs significantly depending on tumor grade,28,29 leading some authors16,20,24-27 to advocate for a core biopsy for grading purposes in such patients.
In conclusion, although immunocytochemistry has now become an indispensable tool in routine cytology for diagnosis, prognostication, and targeted therapies, there are also areas that warrant caution.30,31 This retrospective study describes the utility of FNA as a diagnostic adjunct to medical imaging in the preoperative evaluation and management of patients with PancNETs but at the same time shows the limitations of this technique with regard to grading PancNETs.
1. Conrad C, Fernandez-del
2. Jani N, Khalid A, Kaushik N, et al. EUS-guided FNA diagnosis of pancreatic endocrine tumors: new trends identified. Gastrointest Endosc. 2008;67(1):44-50.
3. Frankel WL. Update on pancreatic endocrine tumors. Arch Pathol Lab Med. 2006;130(7):963-966.
4. Nodit L, McGrath KM, Zahid M, et al. Endoscopic ultrasound-guided fine needle aspirate microsatellite loss analysis and pancreatic endocrine tumor outcome. Clin Gastroenterol Hepatol. 2006;4(12):1474-1478.
5. Reid MD, Bagci P,
6. Fesinmeyer MD, Austin MA, Li CI, De Roos AJ, Bowen DJ. Differences in survival by histologic type of pancreatic cancer. Cancer Epidemiol Biomarkers Prev. 2005;14(7):1766-1773.
7. Tang LH, Gonen M, Hedvat C, Modlin IM, Klimstra DS. Objective quantification of the Ki67 proliferative index in neuroendocrine tumors of the gastroenteropancreatic system: a comparison of digital image analysis with manual methods. Am J Surg Pathol. 2012;36(12):1761-1770.
8. Adsay V. Ki67 labelling index in neuroendocrine tumors of the gastrointestinal and pancreatobiliary tract: to count or not to count is not the question, but rather how to count. Am J Surg Pathol. 2012;36(12):1743-1746.
9. Scarpa A, Mantovani W, Capelli P, et al. Pancreatic endocrine tumors: improved TNM staging and histopathological grading permit a clinically efficient prognostic stratification of patients. Mod Pathol. 2010;23(6):824-833.
10. Fisher SB, Kooby DA. Laparoscopic pancreatectomy for malignancy. JSurg Oncol. 2013;107(1):39-50.
11. DiNorcia J, Lee MK, Reavey PL, et al. One hundred thirty resections for pancreatic neuroendocrine tumor: evaluating the impact of minimally invasive and parenchyma-sparing techniques. J Gastrointest Surg. 2010;14(10):1536- 1546.
12. Pitt SC, Pitt HA, Baker MS, et al. Small pancreatic and periampullary neuroendocrine tumors: resect or enucleate? J Gastrointest Surg. 2009;13(9): 1692-1698.
14. Turner BG, Cizginer S, Agarwal D, Yang J, Pitman MB, Brugge WR. Diagnosis of pancreatic neoplasia with EUS and FNA: a report of accuracy. Gastrointest Endosc. 2010;71(1):91-98.
15. Chang F, Chandra A, Culora G, Mahadeva U, Meenan J, Herbert A. Cytologic diagnosis of pancreatic endocrine tumors by endoscopic ultrasoundguided fine-needle aspiration: a review. Diagn Cytopathol. 2006;34(9):649-658.
16. Levy MJ, Jondal ML, Clain J, Wiersema J. Preliminary experience with an EUS-guided trucut biopsy needle compared with EUS-guided FNA. Gastrointest Endosc. 2003;57(1):101-106.
18. Gu M, Ghafari S, Lin F, Ramzy I. Cytological diagnosis of endocrine tumors of the pancreas by endoscopic ultrasound-guided fine-needle aspiration biopsy. Diagn Cytopathol. 2004;32(4):204-210.
19. Chatzipantelis P, Salla C, Konstantinou P, Karoumpalis I, Sakellariou S, Doumani I. Endoscopic ultrasound-guided fine-needle aspiration cytology of pancreatic neuroendocrine tumors: a study of 48 cases. Cancer Cytopathol. 2008; 114(4):255-262.
20. Wittmann J, Kocjan G, Sgouros SN, Deheragoda M, Pereira SP. Endoscopic ultrasound-guided tissue sampling by combined fine needle aspiration and trucut needle biopsy: a prospective study. Cytopathology. 2006;17(1):27-33.
21. Reid M. Cytology of the pancreas: a practical review for cytopathologists. Surg Pathol Clin. 2011;4(2):651-691.
22. Leiman G. My approach to pancreatic fine needle aspiration. J Clin Pathol. 2007;60(1):43-49.
23. Khan MS,
24. Larghi A, Capurso G, Carnuccio A, et al. Ki-67 grading of nonfunctioning pancreatic neuroendocrine tumors on histologic samples obtained by EUSguided fine-needle tissue acquisition: a prospective study. Gastrointest Endosc. 2012;76(3):570-577.
25. Chatzipantelis P, Konstantinou P, Kaklamanos M, Apostolou G,
26. Hartwig W, Schneider L, Diener MK, Bergmann F, Buchler MW, Werner J. Preoperative tissue diagnosis for tumors of the pancreas. Brit J Surg. 2009;96(1): 5-20.
27. Figueiredo FAF, Giovannini M, Monges G, et al. EUS-FNA predicts 5-year survival in pancreatic endocrine tumors. Gastrointest Endosc. 2009;70(5):907- 914.
28. Yao JC, Phan AT, Chang DZ, et al. Efficacy of RAD001 (everolimus) and octreotide LAR in advanced low- to intermediate-grade neuroendocrine tumors: results of a phase II study. J Clin Oncol. 2008;26(26):4311-4318.
29. Raymond E, Dahan L, Raoul J-L, et al. Sunitinib malate for the treatment of pancreatic neuroendocrine tumors. N Engl J Med. 2011;364(6):501-513.
30. Skoog L, Tani E. Immunocytochemistry: an indispensable technique in routine cytology. Cytopathology. 2011;22(4):215-229.
31. Fowler LJ, Lachar WA. Application of immunohistochemistry to cytology. Arch Pathol Lab Med. 2008;132(3):373-383.
Accepted for publication
From the Departments of Pathology & Laboratory Medicine (Drs Moyana and Hakim) and Radiological Sciences (Dr Shabana) and the Divisions of Radiation Oncology (Drs Kendal and Grimard), Gastroenterology (Dr Chatterjee), Medical Oncology (Drs Maroun and Jonker), and Hepatobiliary Surgery (Dr Mimeault),
The authors have no relevant financial interest in the products or companies described in this article.
Most Popular Stories
- Consumer Prices Edge Up, Surprising Economists
- Stock Market Seesaw Gives Investors Pause
- 'Twilight's' Kristen Stewart Captures Work of Guantánamo Guards in 'Camp X-Ray'
- Weight Loss Claim for Green Coffee Beans Doesn't Hold Water
- Mexico's Pemex Reports Loss
- Holdout Files New Complaint Against Argentina Over Defaulted Debt
- 'Kissing Congressman' Admits Mistake in Reelection Bid
- Steris to Add 100 Jobs in Birmingham
- Beepi Eases Peer-to-peer Auto Sales
- Europe's Bank Test Aims to Strengthen Economy