- Email: [email protected]
Clinical Features and Natural History of Serous Cystic Neoplasm of the Pancreas
Original Paper Pancreatology 2010;10:695–701 DOI: 10.1159/000320694
Received: March 4, 2010 Accepted after revision: August 19, 2010 Published online: January 18, 2011
Clinical Features and Natural History of Serous Cystic Neoplasm of the Pancreas Mitsuharu Fukasawa Hiroyuki Maguchi Kuniyuki Takahashi Akio Katanuma Manabu Osanai Akira Kurita Tamaki Ichiya Takayoshi Tsuchiya Toshifumi Kin Center for Gastroenterology, Teine-Keijinkai Hospital, Sapporo, Japan
Key Words Serous cystic neoplasm Tumor classification Tumor growth
Abstract Aims: To clarify the clinical features and the natural history of serous cystic neoplasm (SCN) of the pancreas. Methods: We retrospectively analyzed data from 30 patients affected by SCN. SCNs were classified as (1) microcystic type, (2) micro- and macrocystic type, and (3) macrocystic type according to the modified WHO classification. Eighteen patients who underwent serial radiographic imaging were identified, and tumor growth rate in these patients was evaluated. Results: The median age was 62 years, and the female:male ratio was 2:1. Twenty-five patients (83%) were asymptomatic and 5 (17%) were symptomatic. The median tumor size was 2.6 cm. Fifteen cases (50%) had the microcystic type, 7 (23%) the micro- and macrocystic type, and 8 (27%) the macrocystic type. Age, gender, symptoms, location or tumor size did not differ significantly among the three subtypes. Eighteen patients were followed up for a median of 58 months. Morphological changes were observed in 3 patients (17%) and enlargement of tumor size in 9 patients (50%) during the follow-up. The growth rate was 0.29 cm per year and doubling time was 3.5 years; these rates did not differ among morphological subtypes or size of tumors. Conclusions: In asymptomatic patients with a clear imaging diagnosis of SCN, non-
© 2011 S. Karger AG, Basel and IAP 1424–3903/10/0106–0695$26.00/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
operative management with a careful follow-up should be recommended. Surgery should be suggested in only symptomatic patients, those with giant tumors (110 cm), rapid growing or when the presence of a potentially malignant tumor cannot be excluded. Copyright © 2011 S. Karger AG, Basel and IAP
Introduction
Serous cystic neoplasms (SCNs) of the pancreas are almost always benign, and are among the common primary pancreatic cystic neoplasms. SCNs constitute 10– 15% of all cystic masses of the pancreas and 1–2% of all primary pancreatic neoplasms [1]. They are typically unifocal and well demarcated, often honeycombed cystic masses. The cysts are loculated, contain mucin-free serous fluid, and are lined by cuboidal or flattened epithelium [2]. The origin and etiology of serous cystadenomas are not completely known. Morphologic and immunohistochemical features favor the intralobular, centroacinar, and ductular cells of the pancreas. The World Health Organization (WHO) classifies them into serous microcystic adenomas and serous oligocystic adenomas [2]. Recently, there have been several case reports of a solid serous adenoma identified as a variant of a serous cystadenoma [3–5]. In addition, a handful of cases of serous cystadenocarcinoma have been reported [6–18]. Because Mitsuharu Fukasawa, MD, PhD Center for Gastroenterology Teine-Keijinkai Hospital 1-12-1-40 Maeda, Teine, Sapporo, Hokkaido 006-0811 (Japan) Tel. +81 11 681 8111, Fax +81 11 685 2998, E-Mail fukasawa3286 @ aol.com
of the rarity and little understanding of the natural history of the SCN, there is still considerable controversy over the treatment of these tumors. Surgical resection is generally performed to relieve symptoms, if the SCN is large, or because of the inability to distinguish an SCN from a mucinous neoplasm, which has a greater malignant potential [19]. Resection of all cystic neoplasms of the pancreas has been recommended by some [20–22], whereas others have advocated more selective approaches based on individual cases [23, 24]. The aims of the present study were to identify the clinical features and evaluate methods of diagnosis of SCNs, to clarify their natural history by reviewing serial radiographs, and finally to propose guidelines for the accurate diagnosis and optimal management of patients with SCNs.
Corresponding subtype of WHO classification
Classification 1 Microcystic type A solid
Microcystic SCN
B honeycomb
2 Micro- and macrocystic type
Oligocystic SCN
3 Macrocystic type
Microcyst <1 cm, macrocyst ≥1 cm
Fig. 1. Our classification of SCNs.
Methods Enrolled in this retroactive study were 30 consecutive patients with the diagnosis of SCN at our institution, the Center for Gastroenterology, Teine-Keijinkai Hospital, Sapporo, between January 1998 and March 2009. Inpatient and outpatient medical records were reviewed in detail. Information regarding demographics, presentation, radiographic and endoscopic evaluation, follow-up after diagnosis, surgical treatment (if any) and pathology were obtained. We have classified SCNs as (1) microcystic type, (2) micro- and macrocystic type, and (3) macrocystic type according to the modified WHO classification. SCNs were defined as microcystic when the cysts within the tumor were smaller than 1 cm; when the cysts were larger than or equal to 1 cm, they were classified as macrocystic. Microcystic SCNs were subdivided into honeycomb and solid types based on morphology. The SCNs with macrocysts were subdivided into micro- and macrocystic (microcystic dominant) and macrocystic (macrocystic dominant) types. A unilocular type was not included because of the impossibility of a preoperative diagnosis (fig. 1). Tumor size was calculated from radiographic images. Maximum diameter was used as the primary metric for tumor size. Tumor volume was determined using the formula /6 ! (d1 ! d2 ! d3) [25]. If fewer than three measurements of diameter were available, the formula was modified accordingly: /6 ! (d12 ! d2) or /6 ! (d3). All patients who were not treated by surgery agreed to followup through close observation. Imaging modalities with computed tomography (CT) and/or magnetic resonance cholangiopancreatography (MRCP) were performed once or twice a year. Serial changes in morphology and the maximum diameter of the tumor were monitored during the observation period. Tumors with an increase or decrease in the maximum diameter 110 mm were defined as enlarged or reduced, respectively. To estimate tumor volume growth, doubling times (DTs) were calculated using the Schwartz equation: DT = [(Ln2)(t2 – t1)]/ln(v2/v1), where t is time and v is volume [26, 27].
696
Pancreatology 2010;10:695–701
Table 1. Comparison of demographic findings among subtypes of
SCN Microcystic Micro- and Macrocystic (n = 15) macrocystic (n = 8) (n = 7)
Males/females Median age, years Symptomatic Location, head/body–tail Median tumor size, mm
5/10 67 2 (13%) 5/10 27
2/5 57 2 (29%) 3/4 33
3/5 52 1 (13%) 4/4 27
Statistical analyses were conducted with the 2 test, Fisher’s test, and the Mann-Whitney U test as appropriate. A p value of !0.05 was considered to indicate a statistically significant difference.
Results
Patient Characteristics The median age of the 30 patients was 62 years (range, 24–92), and the female:male ratio was 2: 1. Overall, 25 (83%) patients were asymptomatic, with the tumor incidentally identified during a routine check-up or during an evaluation of another disease. Five patients (17%) were symptomatic: 2 had abdominal pain, 2 had nonspecific symptoms, and one patient presented with pancreatitis. Five patients (17%) presented in the context of other neoFukasawa et al.
a
c
b
d
e
Fig. 2. A 56-year-old female with macrocystic SCN. US (a), CT (b) and MRCP (c) showing a multiseptated lobulated cystic mass. The lesion was interpreted as branch duct IPMN on these imaging studies. d EUS revealing closely aggregated microcysts in a por-
tion of the cystic tumors. e Endoscopic retrograde cholangiopancreatography showing normal pancreatic duct without communication to cystic lesion. Final diagnosis of this lesion was macrocystic SCN.
plasms, including one patient with hepatocellular carcinoma, one patient with breast cancer, one patient with colon cancer and 2 patients with von Hipple-Lindau disease. Twenty-nine cases (97%) were unifocal, and one case (3%) with von Hipple-Lindau disease had multifocal lesions. Of the 30 SCNs, 12 (40%) were localized in the head of the pancreas and 18 (60%) were in the pancreas body to tail. Median tumor diameter was 2.6 cm (range, 1.0–6.3 cm). Dilation of Wirsung duct was observed in 2 patients (7%) at diagnosis. These 2 patients (7%) underwent surgery and others were closely observed without surgical resection. When SCNs were classified by size and number of cysts, 15 patients (50%) had the microcystic type (all honeycomb type), 7 patients (23%) the microand macro-cystic type, and 8 patients (27%) had the macrocystic type. Comparisons of demographic findings among subjects with the three types of SCN are shown in
table 1. Age, gender, symptoms, location or tumor size did not differ significantly between the three subtypes of SCN.
Natural History of Serous Cystic Neoplasm
Pancreatology 2010;10:695–701
Imaging Assessment Of the 30 patients with an SCN, 25 (83%) were studied with ultrasonography (US), 27 (90%) with CT, 21 (70%) with MRCP, and 28 (93%) with endoscopic US (EUS). Of the 25 tumors studied with US, 11 (44%) were demonstrated to have the typical characteristics (honeycomb pattern with central scar) that led to precise diagnosis of SCN, 8 (32%) were interpreted as intraductal papillary mucinous neoplasm (IPMN), 3 (12%) were a solid mass, and 3 (12%) were not detected. All 27 CT scans demonstrated a cystic mass, 18 (67%) of which were diagnosed as an SCN with the honeycomb pattern or central scar with calcifications, 8 (30%) were interpreted as IPMN, 697
Color version available online
Fig. 3. A 24-year-old female with microand macrocystic SCN. a CT demonstrating a multicystic lesion in the head of pancreas. b EUS showing closely aggregated microcysts in the center of the cystic mass. c Gross findings of a serous cystadenoma involving the head of the pancreas. Note the circumscribed microcystic neoplasm with focal scar. d Microscopic findings showing a microcystic architecture. HE. !100.
a
b
c
d
and 1 (3%) was interpreted as a retention cyst. All 21 MRCP studies showed a high intensity mass on T2weighted images, 14 (67%) of which were diagnosed as an SCN with a honeycomb pattern and 7 (33%) were interpreted as an IPMN. All 28 patients who underwent EUS, including the 7 patients misdiagnosed by US, CT, or MRCP, were revealed by EUS to have closely aggregated microcysts in a portion of the cystic tumors (fig. 2). Operation Of the 30 patients with an SCN, 2 underwent surgery: one patient underwent pancreaticoduodenectomy due to pancreatitis and the other underwent distal pancreatectomy for resection of a large tumor in the pancreas tail. Figure 3 shows a serous cystadenoma involving the head of the pancreas in the patient who underwent the pancreaticoduodenectomy. There were no major perioperative complications. Tumor Growth Twenty-eight of the patients who were diagnosed with an SCN did not undergo surgery. All had results of imaging studies consistent with SCN. Of the 28 patients, 18 698
Pancreatology 2010;10:695–701
Table 2. Clinical outcome of SCN
Clinical outcome Morphology Size of tumor1
Patients No change Changed No change Reduced Enlarged
15 (83%) 3 (17%) 9 (50%) 0 (0%) 9 (50%)
1 Changes in size >10 mm were defined as enlarged or reduced.
(64%) were followed by serial radiography. The interval between the first and last study ranged from 10 to 105 months, with a median of 58 months. Table 2 shows the clinical outcome of eighteen SCNs during long-term follow-up. During the follow-up period, morphologic changes were observed in three tumors (17%); microcystic SCN changed to macrocystic SCN with tumor enlargement in one patient, macrocystic SCN changed to micro- and macrocystic SCN with tumor enlargement in Fukasawa et al.
Table 3. Data on 9 patients with enlarged SCN during follow-up
Age years
Sex
Location
Subtype
Initial size mm
Latest size mm
Change in size mm
Follow-up months
32 47 62 67 68 72 26 42 62
F M M F F F M F F
head body head tail head body body head head
micro micro ] macro micro micro micro micro macro macro macro ] micro
37 17 27 15 37 20 14 42 27
53 43 47 31 50 35 24 68 38
16 26 20 16 13 15 10 26 11
67 102 100 59 91 105 73 56 46
Table 4. Relationship between clinical characteristics and clinical
outcome
Males/females Median age, years Location (head/body–tail) Subtype, 1:2:3 Initial size, mm Follow-up, months
No change (n = 9)
Enlarged (n = 9)
3/6 67 3/6 4:3:2 31 40
3/6 53 5/4 6:0:3 26 72
p value NS NS NS NS NS <0.01
cystic type with enlargement. All cases were asymptomatic and continued with regular follow-up. Table 4 shows a comparison of various clinical factors between SCN patients whose tumor did not change and those whose tumor enlarged. There were no significant differences in age, gender, location, morphological subtype or initial size of the tumor between the two groups. The follow-up duration in SCN patients with an enlarged tumor was significantly longer than in those with no change.
Discussion
one patient, and in the third patient the macrocystic SCN changed to a microcystic SCN through tumor shrinkage from 30 to 25 mm. In 9 cases (50%), the tumor was enlarged, and in the remaining 9 cases (50%) there was no change in size during the follow-up period. The median growth rate in the maximum diameter was 0.29 cm per year; in those tumors showing growth, the median DT was 3.5 years. Dilation of the Wirsung duct with tumor growth was observed in one patient with VHL. This patient continued with follow-up because of multiple lesions in whole pancreas. Data on patients with enlarged SCNs are shown in table 3. Of the 9 cases, there were 6 females and 3 males, with a median age of 62 years (range, 26–78 years) at the time of diagnosis. Five tumors were in the head, three in the body and one in the tail of the pancreas. The morphological subtype at the time of diagnosis was microcystic in 6 cases and macrocystic in 3 patients. Morphology of one microcystic SCN changed to the macrocystic type and that of one macrocystic SCN changed to the micro-
SCNs are being increasingly diagnosed with the more widespread use of imaging techniques and improvements in imaging technology [19, 27]. SCNs are subclassified as serous microcystic adenomas and serous oligocystic adenomas by WHO [2]. However, this WHO classification does not reflect the actual structures of SCNs. A systematic classification of SCNs based on radiological examination is needed. In this report, we present a modified morphological classification of SCNs according to radiological features and analyzed the clinical features of subtypes of the SCN. CT findings suggestive of SCN include a central scar with the ‘honeycomb’ appearance of microcysts, which is found in the more common microcystic type of SCN. However, the rare macrocystic type may be more difficult to differentiate from IPMN or MCN based on CT findings [23, 28]. Although other modalities such as MRI and MRCP may be more useful in differentiating IPMN or MCN from SCN, differentiating the macrocystic type of SCN from other cystic tumors is still difficult. EUS has been proposed as an ideal imaging technique for
Natural History of Serous Cystic Neoplasm
Pancreatology 2010;10:695–701
699
pancreatic cystic lesions [29–31]. Ultrasound can readily characterize cysts, and high-resolution imaging of the pancreas can be achieved through endoscopic means. Accuracy of diagnosis can be improved when EUS is added to other imaging modalities, as it can detect small areas of closely aggregated microcysts. In this study, although 6 (75%) cases with the macrocystic type of SCN were misdiagnosed by US, CT and MRCP, EUS allowed the correct diagnosis in all cases. EUS-guided fine needle aspiration may provide a definitive diagnosis, but sensitivity of diagnosis is generally low due to the small amount of material obtained and may not contain diagnostic material [32–34]. Furthermore, in the case of a mucinous lesion, there remains the disturbing unresolved problem of potential spread of malignant cells [35]. Therefore, aspiration should be reserved only for selected cases [34]. Many SCNs are asymptomatic, and the overwhelming majority of these neoplasms are benign, with the exception of only a handful of case reports [6–18] of serous cystadenocarcinoma. However, the optimal treatment of these neoplasms remains controversial. This can be attributed to the limited understanding of their natural history. Bassi et al. [34] reported that among 100 consecutive cases of SCN, 34 cases were followed up, and 2 underwent surgical resection because of frank tumor growth (12 cm/year), while the remaining 32 were conservatively followed up without any significant increase in the diameter of the lesion. Tseng et al. [23] reported that with a median follow-up of 2 years, the growth rate of SCNs was only 0.6 cm per year; however, the growth rate was significantly greater in tumors 64 cm in size (growth rate of 1.98 cm per year and DT of 0.64 years). These are the only studies investigating the natural growth of SCNs to our knowledge, and the natural history of SCNs over the long term remains unclear. In the present study, half of SCNs showed tumor growth, and the observation period was the only factor associated with the tumor growth. Though the growth rate was rather slow (0.29 cm per year and DT was 3.5 years) in our study, the result indicates that any SCNs have a possibility to grow with time irrespective of tumor subtype, size, or location, suggesting that all SCNs should be followed up regularly. Since George et al. [6] first introduced a malignant variant known as serous cystadenocarcinoma, there have been several case reports [6–18] of SCNs with malignant behavior. Malignant SCN is defined by the presence of metastases to extrapancreatic organs or tissues. Vascular and perineural invasion and local invasion into the duodenum or stomach are not criteria for the diagnosis of a malignant variant [36]. According to the criteria, 9 cases 700
Pancreatology 2010;10:695–701
of malignant SCN have been reported in the literature, with liver metastases in 6 patients and lymph node metastases in 3 patients. Eight of the 9 malignant SCNs (89%) were symptomatic and more than 10 cm in diameter. These findings suggest that SCNs acquire malignant potential according to the extent of growth and that symptomatic or giant (110 cm) SCNs have a high risk of malignant potential. Since all patients were asymptomatic and all SCNs were less than 10 cm in size throughout the observation period in our study, the patients were all under conservative management with careful follow-up. In the past, most SCNs were resected because of the difficulty in distinguishing these tumors from other tumors with malignant potential. However, improvements in imaging technology and a better understanding of the pathological behavior of SCNs have resulted in a reduction in surgically managed cases. Based on data in the present study and in the literature, we propose the following guidelines for the management of SCNs. When the diagnosis of an SCN is confidently determined based on clinical and radiographic evidence, nonoperative management with a careful follow-up should be recommended. Based on the slow growth of these neoplasms observed in this study, we recommend serial imaging at 12-month intervals. Surgery should be suggested in only symptomatic patients, giant tumors (110 cm), rapidly growing or, when the presence of a potentially malignant tumor cannot be excluded.
References
1 Horvath KD, Chabot JA: An aggressive resectional approach to cystic neoplasms of the pancreas. Am J Surg 1999;178:269–274. 2 Cappella C, Solcia E, Kloppel G, et al: Serous cystic neoplasms of the pancreas. Pathology and genetics of tumors of the digestive system. WHO classification of tumors of the digestive system. IARC, Lyon, pp 231–233. 3 Perez-Ordonez B, Naseem A, Lieberman PH, et al: Solid serous adenoma of the pancreas. The solid variant of serous cystadenoma? Am J Surg Pathol 1996; 20:1401–1405. 4 Reese SA, Traverso LW, Jacobs TW, et al: Solid serous adenoma of the pancreas: a rare variant within the family of pancreatic serous cystic neoplasms. Pancreas 2006;33:96– 99. 5 Stern JR, Frankel WL, Ellison EC, et al: Solid serous microcystic adenoma of the pancreas. World J Surg Oncol 2007; 5:26–31. 6 George DH, Murphy F, Michalski R, et al: Serous cystadenocarcinoma of the pancreas: a new entity? Am J Surg Pathol 1989;13:61–66.
Fukasawa et al.
7 Kamei K, Funabiki T, Ochiai M, et al: Multifocal pancreatic serous cystadenoma with atypical cells and focal perineural invasion. Int J Pancreatol 1991; 10:161–172. 8 Yoshimi N, Sugie S, Tanaka T, et al: A rare case of serous cystadenocarcinoma of the pancreas. Cancer 1992;69:2449–2453. 9 Ohta T, Nagakawa T, Ito H, et al: A case of serous cystadenoma of the pancreas with focal malignant changes. Int J Pancreatol 1993; 14:283–289. 10 Widmaier U, Mattfekdt T, Siech M, et al: Serous cystadenocarcinoma of the pancreas. Int J Pancreatol 1996; 20:135–139. 11 Eriguchi N, Aoyagi S, Nakayama T, et al: Serous cystadenocarcinoma of the pancreas with liver metastases. J Hepatobiliary Pancreat Surg 1998;5:467–470. 12 Abe H, Kubota K, Mori M, et al: Serous cystadenoma of the pancreas with invasive growth: benign or malignant? Am J Gastroenterol 1998; 93:1963–1966. 13 Ishikawa T, Nakao A, Nomato S, et al: Immunohistochemical and molecular biological studies of serous cystadenoma of the pancreas. Pancreas 1998; 16:40–44. 14 Wu CM, Fishman EK, Hruban RK, et al: Serous cystic neoplasm involving the pancreas and liver: an unusual clinical entity. Abdom Imaging 1999;24:75–77. 15 Shintaku M, Arimoto A, Sakita N: Serous cystadenocarcinoma of the pancreas. Pathol Int 2005;55:573–580. 16 Strobel O, Z’graggen K, Schmitz-Winnenthal FH, et al: Risk of malignancy in serous cystic neoplasms of the pancreas. Digestion 2003;68:24–33. 17 Matsumoto T, Hirano S, Yoda K, et al: Malignant serous cystic neoplasm of the pancreas: report of a case and review of the literature. J Clin Gastroenterol 2005;39:253– 256.
Natural History of Serous Cystic Neoplasm
18 Friebe V, Keck T, Mattern D, et al: Serous cystadenocarcinoma of the pancreas: management of a rare entity. Pancreas 2005; 31: 182–187. 19 Fernandez-del Castillo C, Targarona J, Thayer SP, et al: Incidental pancreatic cysts: clinicopathologic characteristics and comparison with symptomatic patients. Arch Surg 2003;138:424–427. 20 Galanis C, Zamani A, Cameron JL, et al: Resected serous cystic neoplasms of the pancreas: a review of 158 patients with recommendations for treatment. J Gastrointest Surg 2007;11:820–826. 21 Hung JS, Yang CY, Hu RH, et al: Surgical treatment of pancreatic serous cystadenoma: aggressive for operations but limited resections. Pancreas 2007;35:358–360. 22 Le Borgne J, de Calan L, Partensky C: Cystadenomas and cystadenocarcinomas of the pancreas: a multiinstitutional retrospective study of 398 cases. French Surgical Association. Ann Surg 1999;230:152–161. 23 Tseng JF, Warshaw AL, Sahani DV, et al: Serous cystadenoma of the pancreas: tumor growth rate and recommendations for treatment. Ann Surg 2005; 242:413–421. 24 Zanini F, FantiniL, Casadei R, et al: Serous cystic tumors of the pancreas: when to observe and when to operate: a single-center experience. Dig Surg 2008;25:233–239. 25 Collins VP, Loeffler RK, Tivey H: Observations on growth rates of human tumors. Am J Roentgenol Radium Ther Nucl Med 1956; 76:988–1000. 26 Schwartz M: A biomathematical approach to clinical tumor growth. Cancer 1961; 14: 1272–1294. 27 Megibow AJ, Lombardo FP, Guiarise A, et al: Cystic masses: cross-sectional imaging observations and serial follow-up. Abdom Imaging 2001;26:640–647.
28 Lewandrowski K, Warshaw A, Compton C: Macrocystic serous cystic tumor of the pancreas. A morphologic variant differing from microcystic adenoma. Hum Pathol 1992; 23: 871–875. 29 Anderson MA, Scheiman JM: Nonmucinous cystic pancreatic neoplasms. Gastrointest Endosc Clin N Am 2002; 12:769–779. 30 Ariyama J, Suyama M, Satoh K, et al: Endoscopic ultrasound intraductal ultrasound in the diagnosis of small pancreatic tumors. Abdom Imaging 1998;23:380–386. 31 Brugge WR: Role of endoscopic ultrasound in the diagnosis of cystic lesions of the pancreas. Pancreatology 2001; 1:637–640. 32 Moparty B, Logrono R, Nealon WH, et al: The role of endoscopic ultrasound and endoscopic ultrasound-guided fine needle aspiration in distinguishing pancreatic cystic lesions. Diagn Cytopathol 2007;35:18–25. 33 Brugge WR, Lewandrowski K, Lee-Lewandroski E, et al: Diagnosis of pancreatic cystic neoplasms: a report of the cooperative pancreatic cyst study. Gastroenterology 2004; 126:1330–1336. 34 Bassi C, Salvia R, Molinari E, et al: Management of 100 consecutive case of pancreatic serous cystadenoma: wait for symptoms at imaging or vice versa? World J Surg 2003;27: 319–323. 35 Hirooka Y, Goto H, Itoh A, et al: Case of intraductal papillary mucinous tumor in which endosonography-guided fine needle aspiration biopsy caused dissemination. J Gastroenterol Hepatol 2003; 18:1323–1325. 36 Hruban RH, Klimstra DS, Pitman MB: Atlas of Tumor Pathology. Tumors of the Pancreas. 4th series. Washington, American Institute of Pancreatology, 2006, pp 45–47.
Pancreatology 2010;10:695–701
701
Received: March 4, 2010 Accepted after revision: August 19, 2010 Published online: January 18, 2011
Clinical Features and Natural History of Serous Cystic Neoplasm of the Pancreas Mitsuharu Fukasawa Hiroyuki Maguchi Kuniyuki Takahashi Akio Katanuma Manabu Osanai Akira Kurita Tamaki Ichiya Takayoshi Tsuchiya Toshifumi Kin Center for Gastroenterology, Teine-Keijinkai Hospital, Sapporo, Japan
Key Words Serous cystic neoplasm Tumor classification Tumor growth
Abstract Aims: To clarify the clinical features and the natural history of serous cystic neoplasm (SCN) of the pancreas. Methods: We retrospectively analyzed data from 30 patients affected by SCN. SCNs were classified as (1) microcystic type, (2) micro- and macrocystic type, and (3) macrocystic type according to the modified WHO classification. Eighteen patients who underwent serial radiographic imaging were identified, and tumor growth rate in these patients was evaluated. Results: The median age was 62 years, and the female:male ratio was 2:1. Twenty-five patients (83%) were asymptomatic and 5 (17%) were symptomatic. The median tumor size was 2.6 cm. Fifteen cases (50%) had the microcystic type, 7 (23%) the micro- and macrocystic type, and 8 (27%) the macrocystic type. Age, gender, symptoms, location or tumor size did not differ significantly among the three subtypes. Eighteen patients were followed up for a median of 58 months. Morphological changes were observed in 3 patients (17%) and enlargement of tumor size in 9 patients (50%) during the follow-up. The growth rate was 0.29 cm per year and doubling time was 3.5 years; these rates did not differ among morphological subtypes or size of tumors. Conclusions: In asymptomatic patients with a clear imaging diagnosis of SCN, non-
© 2011 S. Karger AG, Basel and IAP 1424–3903/10/0106–0695$26.00/0 Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com
Accessible online at: www.karger.com/pan
operative management with a careful follow-up should be recommended. Surgery should be suggested in only symptomatic patients, those with giant tumors (110 cm), rapid growing or when the presence of a potentially malignant tumor cannot be excluded. Copyright © 2011 S. Karger AG, Basel and IAP
Introduction
Serous cystic neoplasms (SCNs) of the pancreas are almost always benign, and are among the common primary pancreatic cystic neoplasms. SCNs constitute 10– 15% of all cystic masses of the pancreas and 1–2% of all primary pancreatic neoplasms [1]. They are typically unifocal and well demarcated, often honeycombed cystic masses. The cysts are loculated, contain mucin-free serous fluid, and are lined by cuboidal or flattened epithelium [2]. The origin and etiology of serous cystadenomas are not completely known. Morphologic and immunohistochemical features favor the intralobular, centroacinar, and ductular cells of the pancreas. The World Health Organization (WHO) classifies them into serous microcystic adenomas and serous oligocystic adenomas [2]. Recently, there have been several case reports of a solid serous adenoma identified as a variant of a serous cystadenoma [3–5]. In addition, a handful of cases of serous cystadenocarcinoma have been reported [6–18]. Because Mitsuharu Fukasawa, MD, PhD Center for Gastroenterology Teine-Keijinkai Hospital 1-12-1-40 Maeda, Teine, Sapporo, Hokkaido 006-0811 (Japan) Tel. +81 11 681 8111, Fax +81 11 685 2998, E-Mail fukasawa3286 @ aol.com
of the rarity and little understanding of the natural history of the SCN, there is still considerable controversy over the treatment of these tumors. Surgical resection is generally performed to relieve symptoms, if the SCN is large, or because of the inability to distinguish an SCN from a mucinous neoplasm, which has a greater malignant potential [19]. Resection of all cystic neoplasms of the pancreas has been recommended by some [20–22], whereas others have advocated more selective approaches based on individual cases [23, 24]. The aims of the present study were to identify the clinical features and evaluate methods of diagnosis of SCNs, to clarify their natural history by reviewing serial radiographs, and finally to propose guidelines for the accurate diagnosis and optimal management of patients with SCNs.
Corresponding subtype of WHO classification
Classification 1 Microcystic type A solid
Microcystic SCN
B honeycomb
2 Micro- and macrocystic type
Oligocystic SCN
3 Macrocystic type
Microcyst <1 cm, macrocyst ≥1 cm
Fig. 1. Our classification of SCNs.
Methods Enrolled in this retroactive study were 30 consecutive patients with the diagnosis of SCN at our institution, the Center for Gastroenterology, Teine-Keijinkai Hospital, Sapporo, between January 1998 and March 2009. Inpatient and outpatient medical records were reviewed in detail. Information regarding demographics, presentation, radiographic and endoscopic evaluation, follow-up after diagnosis, surgical treatment (if any) and pathology were obtained. We have classified SCNs as (1) microcystic type, (2) micro- and macrocystic type, and (3) macrocystic type according to the modified WHO classification. SCNs were defined as microcystic when the cysts within the tumor were smaller than 1 cm; when the cysts were larger than or equal to 1 cm, they were classified as macrocystic. Microcystic SCNs were subdivided into honeycomb and solid types based on morphology. The SCNs with macrocysts were subdivided into micro- and macrocystic (microcystic dominant) and macrocystic (macrocystic dominant) types. A unilocular type was not included because of the impossibility of a preoperative diagnosis (fig. 1). Tumor size was calculated from radiographic images. Maximum diameter was used as the primary metric for tumor size. Tumor volume was determined using the formula /6 ! (d1 ! d2 ! d3) [25]. If fewer than three measurements of diameter were available, the formula was modified accordingly: /6 ! (d12 ! d2) or /6 ! (d3). All patients who were not treated by surgery agreed to followup through close observation. Imaging modalities with computed tomography (CT) and/or magnetic resonance cholangiopancreatography (MRCP) were performed once or twice a year. Serial changes in morphology and the maximum diameter of the tumor were monitored during the observation period. Tumors with an increase or decrease in the maximum diameter 110 mm were defined as enlarged or reduced, respectively. To estimate tumor volume growth, doubling times (DTs) were calculated using the Schwartz equation: DT = [(Ln2)(t2 – t1)]/ln(v2/v1), where t is time and v is volume [26, 27].
696
Pancreatology 2010;10:695–701
Table 1. Comparison of demographic findings among subtypes of
SCN Microcystic Micro- and Macrocystic (n = 15) macrocystic (n = 8) (n = 7)
Males/females Median age, years Symptomatic Location, head/body–tail Median tumor size, mm
5/10 67 2 (13%) 5/10 27
2/5 57 2 (29%) 3/4 33
3/5 52 1 (13%) 4/4 27
Statistical analyses were conducted with the 2 test, Fisher’s test, and the Mann-Whitney U test as appropriate. A p value of !0.05 was considered to indicate a statistically significant difference.
Results
Patient Characteristics The median age of the 30 patients was 62 years (range, 24–92), and the female:male ratio was 2: 1. Overall, 25 (83%) patients were asymptomatic, with the tumor incidentally identified during a routine check-up or during an evaluation of another disease. Five patients (17%) were symptomatic: 2 had abdominal pain, 2 had nonspecific symptoms, and one patient presented with pancreatitis. Five patients (17%) presented in the context of other neoFukasawa et al.
a
c
b
d
e
Fig. 2. A 56-year-old female with macrocystic SCN. US (a), CT (b) and MRCP (c) showing a multiseptated lobulated cystic mass. The lesion was interpreted as branch duct IPMN on these imaging studies. d EUS revealing closely aggregated microcysts in a por-
tion of the cystic tumors. e Endoscopic retrograde cholangiopancreatography showing normal pancreatic duct without communication to cystic lesion. Final diagnosis of this lesion was macrocystic SCN.
plasms, including one patient with hepatocellular carcinoma, one patient with breast cancer, one patient with colon cancer and 2 patients with von Hipple-Lindau disease. Twenty-nine cases (97%) were unifocal, and one case (3%) with von Hipple-Lindau disease had multifocal lesions. Of the 30 SCNs, 12 (40%) were localized in the head of the pancreas and 18 (60%) were in the pancreas body to tail. Median tumor diameter was 2.6 cm (range, 1.0–6.3 cm). Dilation of Wirsung duct was observed in 2 patients (7%) at diagnosis. These 2 patients (7%) underwent surgery and others were closely observed without surgical resection. When SCNs were classified by size and number of cysts, 15 patients (50%) had the microcystic type (all honeycomb type), 7 patients (23%) the microand macro-cystic type, and 8 patients (27%) had the macrocystic type. Comparisons of demographic findings among subjects with the three types of SCN are shown in
table 1. Age, gender, symptoms, location or tumor size did not differ significantly between the three subtypes of SCN.
Natural History of Serous Cystic Neoplasm
Pancreatology 2010;10:695–701
Imaging Assessment Of the 30 patients with an SCN, 25 (83%) were studied with ultrasonography (US), 27 (90%) with CT, 21 (70%) with MRCP, and 28 (93%) with endoscopic US (EUS). Of the 25 tumors studied with US, 11 (44%) were demonstrated to have the typical characteristics (honeycomb pattern with central scar) that led to precise diagnosis of SCN, 8 (32%) were interpreted as intraductal papillary mucinous neoplasm (IPMN), 3 (12%) were a solid mass, and 3 (12%) were not detected. All 27 CT scans demonstrated a cystic mass, 18 (67%) of which were diagnosed as an SCN with the honeycomb pattern or central scar with calcifications, 8 (30%) were interpreted as IPMN, 697
Color version available online
Fig. 3. A 24-year-old female with microand macrocystic SCN. a CT demonstrating a multicystic lesion in the head of pancreas. b EUS showing closely aggregated microcysts in the center of the cystic mass. c Gross findings of a serous cystadenoma involving the head of the pancreas. Note the circumscribed microcystic neoplasm with focal scar. d Microscopic findings showing a microcystic architecture. HE. !100.
a
b
c
d
and 1 (3%) was interpreted as a retention cyst. All 21 MRCP studies showed a high intensity mass on T2weighted images, 14 (67%) of which were diagnosed as an SCN with a honeycomb pattern and 7 (33%) were interpreted as an IPMN. All 28 patients who underwent EUS, including the 7 patients misdiagnosed by US, CT, or MRCP, were revealed by EUS to have closely aggregated microcysts in a portion of the cystic tumors (fig. 2). Operation Of the 30 patients with an SCN, 2 underwent surgery: one patient underwent pancreaticoduodenectomy due to pancreatitis and the other underwent distal pancreatectomy for resection of a large tumor in the pancreas tail. Figure 3 shows a serous cystadenoma involving the head of the pancreas in the patient who underwent the pancreaticoduodenectomy. There were no major perioperative complications. Tumor Growth Twenty-eight of the patients who were diagnosed with an SCN did not undergo surgery. All had results of imaging studies consistent with SCN. Of the 28 patients, 18 698
Pancreatology 2010;10:695–701
Table 2. Clinical outcome of SCN
Clinical outcome Morphology Size of tumor1
Patients No change Changed No change Reduced Enlarged
15 (83%) 3 (17%) 9 (50%) 0 (0%) 9 (50%)
1 Changes in size >10 mm were defined as enlarged or reduced.
(64%) were followed by serial radiography. The interval between the first and last study ranged from 10 to 105 months, with a median of 58 months. Table 2 shows the clinical outcome of eighteen SCNs during long-term follow-up. During the follow-up period, morphologic changes were observed in three tumors (17%); microcystic SCN changed to macrocystic SCN with tumor enlargement in one patient, macrocystic SCN changed to micro- and macrocystic SCN with tumor enlargement in Fukasawa et al.
Table 3. Data on 9 patients with enlarged SCN during follow-up
Age years
Sex
Location
Subtype
Initial size mm
Latest size mm
Change in size mm
Follow-up months
32 47 62 67 68 72 26 42 62
F M M F F F M F F
head body head tail head body body head head
micro micro ] macro micro micro micro micro macro macro macro ] micro
37 17 27 15 37 20 14 42 27
53 43 47 31 50 35 24 68 38
16 26 20 16 13 15 10 26 11
67 102 100 59 91 105 73 56 46
Table 4. Relationship between clinical characteristics and clinical
outcome
Males/females Median age, years Location (head/body–tail) Subtype, 1:2:3 Initial size, mm Follow-up, months
No change (n = 9)
Enlarged (n = 9)
3/6 67 3/6 4:3:2 31 40
3/6 53 5/4 6:0:3 26 72
p value NS NS NS NS NS <0.01
cystic type with enlargement. All cases were asymptomatic and continued with regular follow-up. Table 4 shows a comparison of various clinical factors between SCN patients whose tumor did not change and those whose tumor enlarged. There were no significant differences in age, gender, location, morphological subtype or initial size of the tumor between the two groups. The follow-up duration in SCN patients with an enlarged tumor was significantly longer than in those with no change.
Discussion
one patient, and in the third patient the macrocystic SCN changed to a microcystic SCN through tumor shrinkage from 30 to 25 mm. In 9 cases (50%), the tumor was enlarged, and in the remaining 9 cases (50%) there was no change in size during the follow-up period. The median growth rate in the maximum diameter was 0.29 cm per year; in those tumors showing growth, the median DT was 3.5 years. Dilation of the Wirsung duct with tumor growth was observed in one patient with VHL. This patient continued with follow-up because of multiple lesions in whole pancreas. Data on patients with enlarged SCNs are shown in table 3. Of the 9 cases, there were 6 females and 3 males, with a median age of 62 years (range, 26–78 years) at the time of diagnosis. Five tumors were in the head, three in the body and one in the tail of the pancreas. The morphological subtype at the time of diagnosis was microcystic in 6 cases and macrocystic in 3 patients. Morphology of one microcystic SCN changed to the macrocystic type and that of one macrocystic SCN changed to the micro-
SCNs are being increasingly diagnosed with the more widespread use of imaging techniques and improvements in imaging technology [19, 27]. SCNs are subclassified as serous microcystic adenomas and serous oligocystic adenomas by WHO [2]. However, this WHO classification does not reflect the actual structures of SCNs. A systematic classification of SCNs based on radiological examination is needed. In this report, we present a modified morphological classification of SCNs according to radiological features and analyzed the clinical features of subtypes of the SCN. CT findings suggestive of SCN include a central scar with the ‘honeycomb’ appearance of microcysts, which is found in the more common microcystic type of SCN. However, the rare macrocystic type may be more difficult to differentiate from IPMN or MCN based on CT findings [23, 28]. Although other modalities such as MRI and MRCP may be more useful in differentiating IPMN or MCN from SCN, differentiating the macrocystic type of SCN from other cystic tumors is still difficult. EUS has been proposed as an ideal imaging technique for
Natural History of Serous Cystic Neoplasm
Pancreatology 2010;10:695–701
699
pancreatic cystic lesions [29–31]. Ultrasound can readily characterize cysts, and high-resolution imaging of the pancreas can be achieved through endoscopic means. Accuracy of diagnosis can be improved when EUS is added to other imaging modalities, as it can detect small areas of closely aggregated microcysts. In this study, although 6 (75%) cases with the macrocystic type of SCN were misdiagnosed by US, CT and MRCP, EUS allowed the correct diagnosis in all cases. EUS-guided fine needle aspiration may provide a definitive diagnosis, but sensitivity of diagnosis is generally low due to the small amount of material obtained and may not contain diagnostic material [32–34]. Furthermore, in the case of a mucinous lesion, there remains the disturbing unresolved problem of potential spread of malignant cells [35]. Therefore, aspiration should be reserved only for selected cases [34]. Many SCNs are asymptomatic, and the overwhelming majority of these neoplasms are benign, with the exception of only a handful of case reports [6–18] of serous cystadenocarcinoma. However, the optimal treatment of these neoplasms remains controversial. This can be attributed to the limited understanding of their natural history. Bassi et al. [34] reported that among 100 consecutive cases of SCN, 34 cases were followed up, and 2 underwent surgical resection because of frank tumor growth (12 cm/year), while the remaining 32 were conservatively followed up without any significant increase in the diameter of the lesion. Tseng et al. [23] reported that with a median follow-up of 2 years, the growth rate of SCNs was only 0.6 cm per year; however, the growth rate was significantly greater in tumors 64 cm in size (growth rate of 1.98 cm per year and DT of 0.64 years). These are the only studies investigating the natural growth of SCNs to our knowledge, and the natural history of SCNs over the long term remains unclear. In the present study, half of SCNs showed tumor growth, and the observation period was the only factor associated with the tumor growth. Though the growth rate was rather slow (0.29 cm per year and DT was 3.5 years) in our study, the result indicates that any SCNs have a possibility to grow with time irrespective of tumor subtype, size, or location, suggesting that all SCNs should be followed up regularly. Since George et al. [6] first introduced a malignant variant known as serous cystadenocarcinoma, there have been several case reports [6–18] of SCNs with malignant behavior. Malignant SCN is defined by the presence of metastases to extrapancreatic organs or tissues. Vascular and perineural invasion and local invasion into the duodenum or stomach are not criteria for the diagnosis of a malignant variant [36]. According to the criteria, 9 cases 700
Pancreatology 2010;10:695–701
of malignant SCN have been reported in the literature, with liver metastases in 6 patients and lymph node metastases in 3 patients. Eight of the 9 malignant SCNs (89%) were symptomatic and more than 10 cm in diameter. These findings suggest that SCNs acquire malignant potential according to the extent of growth and that symptomatic or giant (110 cm) SCNs have a high risk of malignant potential. Since all patients were asymptomatic and all SCNs were less than 10 cm in size throughout the observation period in our study, the patients were all under conservative management with careful follow-up. In the past, most SCNs were resected because of the difficulty in distinguishing these tumors from other tumors with malignant potential. However, improvements in imaging technology and a better understanding of the pathological behavior of SCNs have resulted in a reduction in surgically managed cases. Based on data in the present study and in the literature, we propose the following guidelines for the management of SCNs. When the diagnosis of an SCN is confidently determined based on clinical and radiographic evidence, nonoperative management with a careful follow-up should be recommended. Based on the slow growth of these neoplasms observed in this study, we recommend serial imaging at 12-month intervals. Surgery should be suggested in only symptomatic patients, giant tumors (110 cm), rapidly growing or, when the presence of a potentially malignant tumor cannot be excluded.
References
1 Horvath KD, Chabot JA: An aggressive resectional approach to cystic neoplasms of the pancreas. Am J Surg 1999;178:269–274. 2 Cappella C, Solcia E, Kloppel G, et al: Serous cystic neoplasms of the pancreas. Pathology and genetics of tumors of the digestive system. WHO classification of tumors of the digestive system. IARC, Lyon, pp 231–233. 3 Perez-Ordonez B, Naseem A, Lieberman PH, et al: Solid serous adenoma of the pancreas. The solid variant of serous cystadenoma? Am J Surg Pathol 1996; 20:1401–1405. 4 Reese SA, Traverso LW, Jacobs TW, et al: Solid serous adenoma of the pancreas: a rare variant within the family of pancreatic serous cystic neoplasms. Pancreas 2006;33:96– 99. 5 Stern JR, Frankel WL, Ellison EC, et al: Solid serous microcystic adenoma of the pancreas. World J Surg Oncol 2007; 5:26–31. 6 George DH, Murphy F, Michalski R, et al: Serous cystadenocarcinoma of the pancreas: a new entity? Am J Surg Pathol 1989;13:61–66.
Fukasawa et al.
7 Kamei K, Funabiki T, Ochiai M, et al: Multifocal pancreatic serous cystadenoma with atypical cells and focal perineural invasion. Int J Pancreatol 1991; 10:161–172. 8 Yoshimi N, Sugie S, Tanaka T, et al: A rare case of serous cystadenocarcinoma of the pancreas. Cancer 1992;69:2449–2453. 9 Ohta T, Nagakawa T, Ito H, et al: A case of serous cystadenoma of the pancreas with focal malignant changes. Int J Pancreatol 1993; 14:283–289. 10 Widmaier U, Mattfekdt T, Siech M, et al: Serous cystadenocarcinoma of the pancreas. Int J Pancreatol 1996; 20:135–139. 11 Eriguchi N, Aoyagi S, Nakayama T, et al: Serous cystadenocarcinoma of the pancreas with liver metastases. J Hepatobiliary Pancreat Surg 1998;5:467–470. 12 Abe H, Kubota K, Mori M, et al: Serous cystadenoma of the pancreas with invasive growth: benign or malignant? Am J Gastroenterol 1998; 93:1963–1966. 13 Ishikawa T, Nakao A, Nomato S, et al: Immunohistochemical and molecular biological studies of serous cystadenoma of the pancreas. Pancreas 1998; 16:40–44. 14 Wu CM, Fishman EK, Hruban RK, et al: Serous cystic neoplasm involving the pancreas and liver: an unusual clinical entity. Abdom Imaging 1999;24:75–77. 15 Shintaku M, Arimoto A, Sakita N: Serous cystadenocarcinoma of the pancreas. Pathol Int 2005;55:573–580. 16 Strobel O, Z’graggen K, Schmitz-Winnenthal FH, et al: Risk of malignancy in serous cystic neoplasms of the pancreas. Digestion 2003;68:24–33. 17 Matsumoto T, Hirano S, Yoda K, et al: Malignant serous cystic neoplasm of the pancreas: report of a case and review of the literature. J Clin Gastroenterol 2005;39:253– 256.
Natural History of Serous Cystic Neoplasm
18 Friebe V, Keck T, Mattern D, et al: Serous cystadenocarcinoma of the pancreas: management of a rare entity. Pancreas 2005; 31: 182–187. 19 Fernandez-del Castillo C, Targarona J, Thayer SP, et al: Incidental pancreatic cysts: clinicopathologic characteristics and comparison with symptomatic patients. Arch Surg 2003;138:424–427. 20 Galanis C, Zamani A, Cameron JL, et al: Resected serous cystic neoplasms of the pancreas: a review of 158 patients with recommendations for treatment. J Gastrointest Surg 2007;11:820–826. 21 Hung JS, Yang CY, Hu RH, et al: Surgical treatment of pancreatic serous cystadenoma: aggressive for operations but limited resections. Pancreas 2007;35:358–360. 22 Le Borgne J, de Calan L, Partensky C: Cystadenomas and cystadenocarcinomas of the pancreas: a multiinstitutional retrospective study of 398 cases. French Surgical Association. Ann Surg 1999;230:152–161. 23 Tseng JF, Warshaw AL, Sahani DV, et al: Serous cystadenoma of the pancreas: tumor growth rate and recommendations for treatment. Ann Surg 2005; 242:413–421. 24 Zanini F, FantiniL, Casadei R, et al: Serous cystic tumors of the pancreas: when to observe and when to operate: a single-center experience. Dig Surg 2008;25:233–239. 25 Collins VP, Loeffler RK, Tivey H: Observations on growth rates of human tumors. Am J Roentgenol Radium Ther Nucl Med 1956; 76:988–1000. 26 Schwartz M: A biomathematical approach to clinical tumor growth. Cancer 1961; 14: 1272–1294. 27 Megibow AJ, Lombardo FP, Guiarise A, et al: Cystic masses: cross-sectional imaging observations and serial follow-up. Abdom Imaging 2001;26:640–647.
28 Lewandrowski K, Warshaw A, Compton C: Macrocystic serous cystic tumor of the pancreas. A morphologic variant differing from microcystic adenoma. Hum Pathol 1992; 23: 871–875. 29 Anderson MA, Scheiman JM: Nonmucinous cystic pancreatic neoplasms. Gastrointest Endosc Clin N Am 2002; 12:769–779. 30 Ariyama J, Suyama M, Satoh K, et al: Endoscopic ultrasound intraductal ultrasound in the diagnosis of small pancreatic tumors. Abdom Imaging 1998;23:380–386. 31 Brugge WR: Role of endoscopic ultrasound in the diagnosis of cystic lesions of the pancreas. Pancreatology 2001; 1:637–640. 32 Moparty B, Logrono R, Nealon WH, et al: The role of endoscopic ultrasound and endoscopic ultrasound-guided fine needle aspiration in distinguishing pancreatic cystic lesions. Diagn Cytopathol 2007;35:18–25. 33 Brugge WR, Lewandrowski K, Lee-Lewandroski E, et al: Diagnosis of pancreatic cystic neoplasms: a report of the cooperative pancreatic cyst study. Gastroenterology 2004; 126:1330–1336. 34 Bassi C, Salvia R, Molinari E, et al: Management of 100 consecutive case of pancreatic serous cystadenoma: wait for symptoms at imaging or vice versa? World J Surg 2003;27: 319–323. 35 Hirooka Y, Goto H, Itoh A, et al: Case of intraductal papillary mucinous tumor in which endosonography-guided fine needle aspiration biopsy caused dissemination. J Gastroenterol Hepatol 2003; 18:1323–1325. 36 Hruban RH, Klimstra DS, Pitman MB: Atlas of Tumor Pathology. Tumors of the Pancreas. 4th series. Washington, American Institute of Pancreatology, 2006, pp 45–47.
Pancreatology 2010;10:695–701
701