Original article
1071
Authors
Won Jae Yoon1, 2, Ebubekir S. Daglilar1, Mari Mino-Kenudson3, Vicente Morales-Oyarvide3, Martha B. Pitman3, William R. Brugge4
Institutions
Institutions are listed at the end of article.
submitted 20. January 2014 accepted after revision 30. June 2014
Background and study aims: Intraductal papillary mucinous neoplasm (IPMN) consists of four epithelial subtypes. There are limited data on the endoscopic ultrasound (EUS) findings and/or cyst fluid analysis of the epithelial subtypes. The objective of this study was to determine whether there are differences in EUS and cyst fluid characteristics (carcinoembryonic antigen [CEA] concentration and cytology) among the subtypes. Patients and methods: The study cohort consisted of 85 patients (median age 68 years, 40 men) with resected and histologically confirmed branchduct or mixed-type IPMNs who underwent preoperative EUS-guided fine-needle aspiration between 1999 and 2010 for the evaluation of pancreatic cysts. EUS and cyst fluid characteristics were analyzed retrospectively and correlated with the subtypes. Results: The numbers of evaluated cystic lesions were 1 in 79 patients, 2 in 5 patients, and 3 in 1
patient. Of 92 IPMNs analyzed, gastric-type IPMNs were the most common (n = 68, 73.9 %), followed by intestinal (n = 17, 18.5 %), oncocytic (n = 5, 5.4 %), and pancreatobiliary subtypes (n = 2, 2.2 %). Gastric-type IPMNs were significantly smaller (cutoff 30 mm; P = 0.002), and less likely than other subtypes to have a mass lesion or mural nodule (P = 0.046) on EUS. Cyst fluid CEA concentration varied among the subtypes (median concentrations for gastric, intestinal, oncocytic, and pancreatobiliary types 619.8, 83.0, 5.1, and 270.0 ng/mL, respectively; P = 0.012). The presence of neoplastic epithelial cells (P = 0.624) and extracellular mucin (P = 0.208) on cytology had no association with subtypes. Conclusions: Gastric-type IPMNs, the most common subtype, are characterized by high concentrations of cyst fluid CEA, small cyst diameter, and low risk EUS imaging features.
Introduction
teins, IPMNs are classified into four epithelial subtypes: gastric, intestinal, oncocytic, and pancreatobiliary [8]. The epithelial subtypes are reported to be significantly associated with a variety of clinicopathological features such as sex, age, histological grade, differential involvement of the pancreatic duct system, invasive phenotypes, stage, recurrence, and patient survival [9 – 12]. Gastric-type IPMNs are reported to have a more favorable prognosis compared with other epithelial subtypes [9]. Endoscopic ultrasound (EUS) is being increasingly utilized in the diagnosis of IPMN. EUS provides detailed morphological analysis of PCLs. EUSguided fine-needle aspiration (EUS-FNA) enables cyst fluid analysis such as carcinoembryonic antigen (CEA) and cytology [13, 14]. Cyst fluid CEA concentration has been reported to be the most accurate marker to differentiate mucinous (i. e. IPMN and mucinous cystic neoplasm) from nonmucinous PCLs [13, 15].
Bibliography DOI http://dx.doi.org/ 10.1055/s-0034-1377629 Published online: 10.9.2014 Endoscopy 2014; 46: 1071– 1077 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0013-726X Corresponding author William R. Brugge, MD Gastrointestinal Unit Harvard Medical School Massachusetts General Hospital 55 Fruit Street Boston, MA 02114 United States Fax: +1-617-724-5997 [email protected]
!
Intraductal papillary mucinous neoplasm of the pancreas (IPMN) is defined as a grossly visible, intraductal, epithelial neoplasm composed of mucin-producing cells, arising in the main pancreatic duct and/or its branches [1]. The incidence of IPMN is reported to be approximately 5 % among pancreatic resection specimens and 20 % – 40 % among all pancreatic cystic neoplasms [1 – 4]. However, these figures represent surgical data, and therefore the true incidence of IPMNs is most likely to be higher. Indeed, the prevalence of pancreatic cystic lesions (PCLs) ranges between 1.2 % and 19.6 % according to image-based studies [5 – 7], and a significant proportion of these lesions are probably IPMNs. Recently, morphological variations in the epithelial cell types of IPMNs have been recognized. Based on histomorphological features and immunohistochemical reactivity for mucin glycopro-
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Characterization of epithelial subtypes of intraductal papillary mucinous neoplasm of the pancreas with endoscopic ultrasound and cyst fluid analysis
Original article
Determination of the epithelial subtypes of IPMN is only possible through the histological evaluation of the resected specimens. The use of pancreatic juice obtained during endoscopic retrograde cholangiopancreatography for cytology or DNA methylation analysis of mucin expression for the prediction of epithelial subtypes of IPMN has been reported [16, 17]. However, to the best of our knowledge, there are no available data regarding EUS findings and/or cyst fluid characteristics of the epithelial subtypes. Determination of such association may aid in the prediction of the clinicopathological behavior of the lesion, and furthermore, in the planning of treatment. The aims of this study were to determine 1) EUS imaging features, 2) cyst fluid CEA concentration, and 3) the cytological characteristics of the epithelial subtypes of IPMN. It was hypothesized that such characteristics might be associated with particular epithelial subtypes.
Patients and methods !
Study population The study was approved by the Partners Human Research Committee Institutional Review Board. A prospective database was retrospectively analyzed to identify all patients who underwent EUS-FNA to evaluate PCLs at Massachusetts General Hospital. The study cohort consisted of patients who underwent surgery and were ultimately diagnosed with IPMNs by histopathological analysis between 1999 and 2010. The decision to undergo a surgical resection was made based on the 2006 international consensus guideline in the majority of cases [18]. These parameters (presumed main-duct type or mixed-type IPMN, branch-duct IPMN [BD-IPMN] > 3 cm in diameter, presence of a mass lesion or mural nodule, positive cytology, and/or presence of symptoms) had been followed for a few years at Massachusetts General Hospital before the guidelines were published. Other patients were referred for resection based on the recommendation of their physician or surgeon. The following parameters were analyzed: sex, age, EUS and FNA characteristics (location, size in long diameter, presence of septation, mass lesion, mural nodule, and thick cyst wall, and cyst fluid viscosity as reported by endosonographer), cyst fluid CEA concentration, cyst fluid cytology results (presence of neoplastic epithelial cells or extracellular mucin, and cytological grade of dysplasia), and surgical histology results (epithelial subtype, type of duct involvement, and grade of dysplasia). For patients with multiple PCLs evaluated with EUS-FNA, each lesion was independently analyzed. In such patients, the location and the morphological features including the size of the cyst were carefully compared between the EUS findings and surgical specimens. Patients were excluded from the study if the time interval between EUS-FNA and surgery was more than 1 year, and in cases in which EUS showed distinctive cystic lesions but histology confirmed main-duct type IPMN (MD-IPMN). These criteria were applied so that the results of EUS-FNA would reflect the cyst fluid characteristics in a timely manner, and to exclude the non-neoplastic branch-duct dilation secondary to MD-IPMNs.
EUS-FNA procedure All EUS-FNAs were performed as previously described [13, 15]. In brief, EUS was performed using a curvilinear echoendoscope by three experienced endosonographers. The PCL was aspirated under EUS guidance. Vascular structures were identified and avoided using Doppler imaging. PCLs in the pancreatic head or
uncinate process were accessed via a transduodenal approach; those in the body or tail were aspirated via a transgastric approach. To avoid infection of the PCL, intravenous antibiotics were administered during the procedure; an oral quinolone was administered for 2 – 3 days after the procedure in selected cases. When adequate amount of cyst fluid was aspirated, the sample was sent for both CEA concentration and cytology. When limited amount of cyst fluid was aspirated, the sample was triaged according to the clinical question. If the primary question was whether the PCL was mucinous or nonmucinous, the priority was given to CEA concentration. If the primary question was whether the PCL was malignant or benign, the priority was given to cytology [19].
Cyst fluid CEA concentration Cyst fluid CEA concentration was measured as previously described [13, 14]. To summarize, undiluted cyst fluid obtained from EUS-FNA was analyzed for CEA by a commercially available radioimmunoassay kit (Abbott Laboratory, Diagnostics Division, Abbott Park, Illinois, USA). When CEA concentration of a PCL was reported as a range, the lowest value of the range was considered to be the CEA level. For example, when cyst fluid CEA concentration was reported to be > 500 ng/mL, the cyst fluid CEA concentration was considered to be 500 ng/mL.
Cytological examination of the cyst fluid Cytology specimens consisted of direct smears, cytospins or liquid-based preparations. They were routinely stained with Papanicolaou stain. Rarely, hematoxylin and eosin stains were used for rapid interpretation. Mucicarmine and alcian blue stains were used to identify mucin in cyst aspirates with an adequate amount of fluid. In the cytology reports of 24 PCLs, the results of mucin stains were not available; these lesions were considered to have no extracellular mucin present. Epithelial cells in the cyst fluid were evaluated for cytological atypia. In brief, the cytology slides were re-reviewed by a cytopathologist and the epithelial cells were classified in a two-tiered system of low grade or high grade epithelial atypia. Low grade epithelial atypia represented cells with low to intermediate grade dysplasia or cells indeterminate for low grade dysplasia vs. gastrointestinal contamination. High grade epithelial atypia represented cells with either high grade dysplasia or invasive carcinoma [20, 21].
Histological examination of IPMN The pathology slides from the surgical specimen were reviewed by a single pathologist. Epithelial subtype, grade of dysplasia, and type of duct involvement by IPMN were evaluated. In patients with multiple cysts resected, all the specimens were reviewed. The epithelial subtype was determined on the basis of histomorphology and mucin glycoprotein immunostaining [8]. The grade of dysplasia was reported based on the highest degree of the dysplasia identified. The type of duct involvement was determined based on the results of pathology and imaging studies [10]. If PCLs had a pure single epithelial subtype, it was recorded as such. In cases exhibiting heterogeneous epithelium, the predominant type of the highest grade was recorded as the subtype and used for analysis [10]. Only invasive carcinoma was considered to be malignant, as stated in both the World Health Organization classification and the 2012 international consensus guideline [1, 22].
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
1072
Original article
Histological characteristics of the PCLs
Patient characteristics.
Number of patients
85
Sex, male/female, n
40/45
Age, median (range), years
68 (39 – 83)
Cystic lesions evaluated, n
92
Patients with lesions, n (%) 1 cystic lesion
79 (92.9)
2 cystic lesions
5 (5.9)
3 cystic lesions
1 (1.2) 53 (4 – 246)
Time between EUS-FNA and surgery, median (range), days EUS-FNA, endoscopic ultrasound-guided fine-needle aspiration.
Statistical analysis All continuous variables are reported as the median (range). For categorical data, the chi-squared test or the Fisher’s exact test was performed, where appropriate. For comparison of cyst fluid CEA concentration, the Kruskal – Wallis test was performed. A two-sided P value of < 0.05 was considered to be statistically significant. STATA 12.1 (StataCorp LP, College Station, Texas, USA) was used for all statistical analyses.
Of the 85 patients, 62 had gastric, 17 had intestinal, 4 had oncocytic, and 2 had the pancreatobiliary subtype IPMN. A PCL with mixed epithelial subtype was identified in four patients. The mixed subtypes were pancreatobiliary-oncocytic (n = 1), pancreatobiliary-gastric (n = 1), gastric-intestinal (n = 1), and intestinal-gastric (n = 1). They were considered to be pancreatobiliary, pancreatobiliary, gastric, and intestinal subtypes, respectively, based on the grade and predominance of the subtypes within the tumor. BD-IPMNs were diagnosed in 40 gastric, 8 intestinal, 2 oncocytic, and no pancreatobiliary IPMN patients; the other patients (22 gastric, 9 intestinal, 2 oncocytic, and 2 pancreatobiliary IPMN patients) had mixed-type IPMNs. In all but one patient with multiple PCLs aspirated, the cysts were of the same grade of dysplasia and epithelial subtype. In one patient who had two cysts aspirated, pathology revealed low grade dysplasia in one cyst and IPMN with an associated invasive carcinoma in the other. Epithelial subtypes were significantly associated with the grade of dysplasia (P < 0.001). Subsequent comparison showed that gastric-type IPMNs were significantly associated with noninvasive histology compared with other types of IPMNs (92.6 % for gastric-type vs. 66.7 % for nongastric-type IPMNs; P = 0.004). The " Table 2. histological characteristics are summarized in ●
EUS features Results !
Patient characteristics A total of 1060 patients underwent EUS-FNA for the evaluation of PCLs during the study period. A total of 328 patients underwent resection of IPMNs during the same period. A total of 93 patients with IPMNs who underwent preoperative EUS-FNA of distinctive cystic lesions were identified. Eight patients were excluded from the analysis: five had a time interval between EUS-FNA and surgery of more than 1 year, and three had MD-IPMN on histology. Thus, a total of 85 patients (40 male) with 92 distinctive PCLs were evaluated. The median age at diagnosis was 68 years. The numbers of evaluated PCLs were 1 in 79 patients, 2 in 5 patients, and 3 in 1 patient. The median time interval between EUS-FNA " Table 1). and surgery was 53 days (●
Table 2
The epithelial subtype was associated with the presence of a mass lesion or mural nodule (P = 0.012) and size of the PCL (P = 0.003). Subsequent comparison showed that gastric-type IPMNs were significantly less likely to demonstrate a mass lesion or mural nodule on EUS than other subtypes of IPMNs (17.6 % for gastric-type vs. 37.5 % for nongastric type; P = 0.046). In addition, gastric-type IPMNs were less likely to be ≥ 30 mm in size than other subtypes of IPMNs (21.5 % for gastric-type vs. 57.1 % for nongastric type; P = 0.002). However, the epithelial subtype was not associated with the location of the PCL, presence of a thick cyst wall, septation, or viscous fluid on FNA " Table 3). (●
Histological characteristics of intraductal papillary mucinous neoplasm.
Epithelial subtype
Total
Gastric
Intestinal
Oncocytic
Pancreatobiliary
Patients, n (%)
62 (72.9)
17 (20.0)
4 (4.7)
2 (2.4)
85 (100.0)
Cystic lesions, n (%)
68 (73.9)
17 (18.5)
5 (5.4)
2 (2.2)
92 (100.0)
Branch duct
46 (67.6)
8 (47.1)
2 (40.0)
0 (0)
56 (60.9)
Mixed type
22 (32.4)
9 (52.9)
3 (60.0)
2 (100.0)
36 (39.1)
Low
26 (38.2)
0 (0)
0 (0)
0 (0)
26 (28.3)
Intermediate
39 (42.4)
Macroscopic type, n (%) 2
P value1
0.066
Grade of dysplasia, n (%) 2,3
< 0.001 32 (47.0)
7 (41.2)
0 (0)
0 (0)
High
5 (7.4)
6 (35.3)
2 (40.0)
1 (50.0)
14 (15.2)
Invasive
5 (7.4)
4 (23.5)
3 (60.0)
1 (50.0)
13 (14.1)
IPMN, intraductal papillary mucinous neoplasm. 1 Freeman–Halton extension of the Fisher’s exact test of the four epithelial subtypes. 2 Number and percentage of cystic lesions. 3 Gastric-type IPMNs were significantly associated with noninvasive histology compared with other subtypes of IPMNs (92.6 % gastric-type vs. 66.7 % for nongastric-type IPMNs; P = 0.004, Fisher’s exact test).
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Table 1
1073
Original article
Endoscopic ultrasound characteristics of cystic lesions according to epithelial subtype of intraductal papillary mucinous neoplasm.
Epithelial subtype Gastric Mass lesion/mural nodule, n (%)
P value1
Total
Intestinal
Oncocytic
Pancreatobiliary
2
0.012
Yes
12 (17.6)
4 (23.5)
4 (80.0)
1 (50.0)
21 (22.8)
No
56 (82.4)
13 (76.5)
1 (20.0)
1 (50.0)
71 (77.2)
≥ 30 mm
14 (21.5)
8 (50.0)
3 (100.0)
1 (50.0)
26 (30.2)
< 30 mm
51 (78.5)
8 (50.0)
0 (0.0)
1 (50.0)
60 (69.8)
Head
39 (57.4)
11 (64.7)
3 (60.0)
1 (50.0)
54 (58.7)
Body-tail
29 (42.6)
6 (35.3)
2 (40.0)
1 (50.0)
38 (41.3)
Yes
11 (16.2)
4 (23.5)
2 (40.0)
0 (0.0)
17 (18.5)
No
57 (83.8)
13 (76.5)
3 (60.0)
2 (100.0)
75 (81.5)
Yes
38 (55.9)
9 (52.9)
4 (80.0)
1 (50.0)
52 (56.5)
No
30 (44.1)
8 (47.1)
1 (20.0)
1 (50.0)
40 (43.5)
Yes
38 (82.6)
14 (100.0)
2 (100.0)
2 (100.0)
56 (87.5)
No
8 (17.4)
0 (0.0)
0 (0.0)
Size, n (%) 3,4
0.003
Location, n (%)
0.962
Thick cyst wall, n (%)
0.395
Septation, n (%)
0.796
Viscous fluid, n (%) 5
0.407 0 (0.0)
8 (12.5)
IPMN, intraductal papillary mucinous neoplasm. 1 Freeman–Halton extension of the Fisher’s exact test of the four epithelial subtypes. 2 Gastric-type IPMNs were significantly less likely to demonstrate a mass lesion or mural nodule on EUS than other subtypes of IPMNs (P = 0.046, chi-squared test). 3 Results of 86 cystic lesions with size reported. 4 Gastric-type IPMNs were less likely to be ≥ 30 mm in size than other subtypes of IPMNs (P = 0.002, chi-squared test). 5 Results of 64 cystic lesions with fluid viscosity reported.
Cyst fluid CEA concentration Cyst fluid CEA concentrations were reported in 60 PCLs (49 gastric, 6 intestinal, 4 oncocytic, and 1 pancreatobiliary). The epithelial subtype was significantly associated with cyst fluid CEA concentration (P = 0.012, Kruskal–Wallis test), with the gastric-type IPMN apparently having the highest median concentration " Fig. 1, ● " Table 4). However, the cyst fluid CEA concentration (● was not significantly associated with the grade of dysplasia of " Table 4). IPMNs (P = 0.072) (●
Cyst fluid cytology Cyst fluid cytology was performed in 76 PCLs. Neoplastic epithelial cells were identified in 28 lesions (36.8 %). High grade epithelial atypia was present in 21 of 28 lesions. Extracellular mucin was present in 42 lesions (55.3 %). When the presence of mucin was defined as viscous fluid as described by endosonographer or presence of extracellular mucin on cytological specimen, mucin was present in 72 lesions (78.3 %). The presence of neoplastic epithelial cells or extracellular mucin was not associated with epithelial subtype. The high grade epithelial atypia features showed a tendency to be associated with epithelial subtype (P = " Table 5. 0.092). The cytology results are summarized in ●
Discussion !
The purpose of this study was to analyze the EUS-FNA findings of the epithelial subtypes of IPMN. Analysis of 92 PCLs showed that the gastric-type IPMN was positively associated with high concentrations of cyst fluid CEA and negatively associated with high risk EUS imaging features of mass lesion or mural nodule and large cyst size. Cytological features of cells exfoliated into the cyst fluid were nonspecific to the epithelial subtype.
100000
10000
1000
100
10
1 Gastric
Intestinal
Oncocytic
Pancreatobiliary
Fig. 1 Cyst fluid carcinoembryonic antigen (CEA) concentrations (log scale) according to epithelial subtypes of intraductal papillary mucinous neoplasms (n = 60). Epithelial subtype was significantly associated with cyst fluid CEA concentration (P = 0.012, Kruskal–Wallis test).
IPMNs have been known to have morphological variations that are associated with clinicopathological outcomes [23]. A consensus classification for distinguishing the four epithelial subtypes of IPMN was drawn in 2005, and is based on the morphological features and the immunohistochemical reactivity for mucin glycoproteins [8]. This classification was validated in recent studies [9, 10]. In a large-scale study of 283 cases, the epithelial subtypes were associated with sex, age, duct dilation, detection of mural nodule on imaging, histological grade, pancreatic duct involvement, invasive phenotype, stage, recurrence, and patient survival [9]. Moreover, Mino-Kenudson et al. reported that the prognosis
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Table 3
CEA concentration (ng/mL)
1074
Original article
Table 6 Characteristics of epithelial subtypes of intraductal papillary mucinous neoplasm.
Table 4 Comparison of cyst fluid carcinoembryonic antigen concentrations according to epithelial subtype and grade of dysplasia of intraductal papillary mucinous neoplasm.
Epithelial subtype
0.012 619.8 (22.3 – 50000.0)
Intestinal (n = 6)
83.0 (11.1 – 5165.0)
Oncocytic (n = 4)
5.1 (3.4 – 358.8)
Pancreatobiliary (n = 1)
270.0
Low (n = 23)
619.8 (26.7 – 50000.0)
Intermediate (n = 24)
343.4 (11.1 – 19704.0)
High (n = 6)
185.8 (5.6 – 1872.0)
Invasive (n = 7)
98.4 (3.4 – 5165.0)
Mucin glyco-
Correlated
protein staining
invasive IPMN
1
0.072 1
Grade of dysplasia
Architecture
subtype
Gastric
Thick finger-like papillae
MUC5AC
Tubular adenocarcinoma
Intestinal
Villous papillae
MUC2, MUC5AC
Colloid carcinoma
Oncocytic
Thick branching complex papillae
MUC1, MUC5AC
Oncocytic carcinoma
Pancreatobiliary
Thin branching complex papillae
MUC1, MUC5AC
Tubular adenocarcinoma
IPMN, intraductal papillary mucinous neoplasm.
CEA, carcinoembryonic antigen. 1 Kruskal–Wallis test.
of invasive IPMN was shown to be dependent upon the histology of invasive components and precursor epithelial subtypes [10]. In addition, the histology of invasive components of IPMN was closely associated with the precursor epithelial subtypes. Colloid and oncocytic carcinomas originated only from intestinal and oncocytic IPMN types, respectively, and were frequently of the main-duct type; tubular adenocarcinomas primarily arose from the gastric-type IPMN, and were often associated with BD-IPMN. The colloid and oncocytic types of invasive IPMNs were more indolent compared with tubular types. The authors concluded that epithelial subtypes of precursor IPMN correlate with the histological subtype of invasive carcinoma, and ultimately with prognosis [10]. It is generally accepted that the gastric-type IPMN has a favorable outcome. However, when a carcinoma does develop in gastric-type IPMN, it is usually of the tubular type and demonstrates an aggressive behavior [22]. A brief summary of the fea" Table 6. tures of epithelial subtypes is presented in ● In the current study, gastric-type IPMNs were less likely to exhibit a mass lesion or mural nodule on EUS, and were smaller than other subtypes of IPMNs. This is analogous to the association between the epithelial subtype and image findings reported by Furukawa et al., where the intestinal-type IPMNs were larger than gastric-type IPMNs, and oncocytic and pancreatobiliary subtypes were associated with detection of mural nodules [9].
Table 5
Interestingly, cyst fluid CEA concentration varied significantly with the epithelial subtypes of IPMN. However, there was no statistically significant association between the cyst fluid CEA concentration and the grade of dysplasia. For IPMNs, the ability of cyst fluid CEA to differentiate malignant from benign PCLs has been controversial. Although a recent report indicated that a high CEA concentration in the pancreatic juice is associated with malignancy in BD-IPMNs [24], these results are difficult to be applied directly to the results of EUSFNA, as there are issues of differences in the specimen (cyst fluid vs. pancreatic juice) and the sampling method (EUS-FNA vs. pancreatic juice aspiration during endoscopic retrograde pancreatography). A study of 41 IPMN patients who underwent preoperative EUS-FNA reported that a cyst fluid CEA concentration > 200 ng/mL had sensitivity, specificity, positive predictive value, and negative predictive value of 90 %, 71 %, 50 %, and 96 %, respectively, for the diagnosis of malignant IPMN [25]. However, these findings were not reproduced in subsequent studies. In a report of 72 IPMN patients, the cutoff cyst fluid CEA concentration of 200 ng/mL demonstrated sensitivity, specificity, positive predictive value, and negative predictive value of 47 %, 40 %, 20 %, and 70 %, respectively [26]. Another study by Kucera et al. involving 47 IPMN patients demonstrated that the cyst fluid CEA concentration had no association with grades of dysplasia [27]. Results
Comparison of cytology characteristics according to epithelial subtype of intraductal papillary mucinous neoplasm. Results of 76 cystic lesions.
Epithelial subtype
Cystic lesions, n
Total
Gastric
Intestinal
Oncocytic
Pancreatobiliary
54
15
5
2
76
Neoplastic epithelial cells, n (%)
P value1
0.624
Yes
22 (40.7)
4 (26.7)
1 (20.0)
1 (50.0)
28 (36.8)
No
32 (59.3)
11 (73.3)
4 (80.0)
1 (50.0)
48 (63.2)
High grade epithelial atypia 3
18 (81.8)
1 (25.0)
1 (100.0)
1 (100.0)
21 (75.0)
Low grade epithelial atypia 4
4 (18.2)
3 (75.0)
0 (0.0)
0 (0.0)
7 (25.0)
Yes
28 (51.9)
11 (73.3)
3 (60.0)
0 (0.0)
42 (55.3)
No
26 (48.1)
4 (26.7)
2 (40.0)
2 (100.0)
34 (44.7)
Cytological atypia, n (%) 2
0.092
Extracellular mucin, n (%)
0.208
1
Freeman–Halton extension of the Fisher’s exact test of the four epithelial subtypes. Results from 28 cystic lesions with neoplastic epithelial cells present in the cyst fluid. High grade epithelial atypia represents cells with either high grade dysplasia or invasive carcinoma. 4 Low grade epithelial atypia represents cells with low to intermediate grade dysplasia or cells indeterminate for low grade dysplasia vs. gastrointestinal contamination. 2 3
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Gastric (n = 49)
Epithelial
P value
CEA, median (range), ng/mL
1075
Original article
from the current study suggest that the cyst fluid CEA concentration may be correlated with epithelial subtypes rather than grade of dysplasia. The mechanisms of the difference in the cyst fluid CEA according to the epithelial subtype of IPMN remain to be elucidated. One hypothesis is that the quantitative expression of CEA would be different according to epithelial subtypes. Although CEA is most well known to be expressed in colorectal cancer tissues that share the immunophenotype with intestinal-type IPMNs [28], it is also reported to be expressed in gastric cancer and precursor lesions of gastric cancer [29], and pancreatic ductal adenocarcinoma [30]. Thus, each epithelial subtype of IPMN might have the potential to express CEA. There might be a difference in the quantity of CEA expression according to epithelial subtypes, which may result in different cyst fluid CEA concentrations. Another hypothesis would be the difference in the polarized distribution of CEA according to epithelial subtypes. In colorectal cancer, CEA is reported to be distributed along the luminal border in well-differentiated cancers, whereas it is distributed over the entire cell surface in a diffuse pattern in poorly differentiated cancers [31]. The distribution of CEA is correlated with the plasma CEA level in colorectal cancer. Basolateral distribution of CEA was associated with high plasma CEA level [32]. In IPMNs, luminal distribution of CEA might result in high cyst fluid CEA concentration, whereas basolateral distribution of CEA might result in low concentration of cyst fluid CEA. The presence of neoplastic epithelial cells and extracellular mucin on cytology of cyst fluid was not associated with epithelial subtypes of IPMN. Cytological examination of the cyst fluid mainly evaluates the neoplastic epithelial cells shed into the cyst. Thus, one may expect that malignant lesions would shed more neoplastic epithelial cells than benign ones [19]. Given the fact that gastric-type IPMNs usually exhibit low to intermediate grade dysplasia, and gastric-type IPMNs may have an abundant flat cyst lining, as opposed to other subtypes which predominantly demonstrate papillary architecture [8], it was expected that the presence of neoplastic epithelial cells in cyst fluid would be more frequent in nongastric-type IPMNs. However, no such association was found. Contrary to expectations, although statistically insignificant, there was a trend for gastric-type IPMNs to have neoplastic cells in the FNA specimens more frequently than other epithelial subtypes. One suggested hypothesis to explain this observation is that the epithelium of gastric-type IPMN might shed more neoplastic cells into the cyst fluid. Another hypothesis is that the detached gastric-type neoplastic cells might maintain morphological integrity longer than nongastric subtypes. This study has limitations, including the retrospective and singlecenter nature of the study. In addition, only the lesions that had been surgically resected could be included in the study, resulting in an inevitable selection bias. Reflecting the proportion of the epithelial subtypes of IPMN, the numbers of patients with oncocytic and pancreatobiliary subtypes were small. The proportion of gastric-type IPMN was somewhat higher than in other studies. This may be partially explained by the high proportion of BDIPMNs in the study cohort. One might argue that the MD-IPMN cases should have been included in the analysis. However, the results of Kucera et al. [27] and Hirono et al. [24] suggest that the specimen for fluid analysis should be aspirated from the primary PCL with neoplastic epithelial lining. The features of non-neoplastic branch duct dilation secondary to MD-IPMNs may not be representative of the main neoplasm. To the best of our knowl-
edge, this is the first study to determine the EUS and cyst fluid findings of the epithelial subtypes of IPMN. In conclusion, the cyst fluid CEA concentration seems to be associated with the epithelial subtypes of IPMN; it seems to have a tendency toward negative correlation with the grade of dysplasia. Gastric-type IPMNs, which are strongly associated with low histological grade and are the most common type, are characterized by high concentrations of cyst fluid CEA, small cyst diameter, and low risk EUS imaging features. These features may aid in the prediction of the epithelial subtypes of IPMN and provide valuable information for selection of treatment strategy. Competing interests: Professor Brugge has received research funding from RedPath and Asuragen. Institutions 1 Gastrointestinal Unit, Massachusetts General Hospital, Boston, Massachusetts, United States 2 Department of Internal Medicine, Inje University Seoul Paik Hospital, Inje University College of Medicine, Seoul, Republic of Korea 3 Department of Pathology, Massachusetts General Hospital, Boston, Massachusetts, United States 4 Gastrointestinal Unit, Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts, United States
Acknowledgments !
This work was conducted with support from Harvard Catalyst, The Harvard Clinical and Translational Science Center (National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health Award 8UL1TR000170-05 and financial contributions from Harvard University and its affiliated academic healthcare centers). The content is solely the responsibility of the authors and does not necessarily represent the official views of Harvard Catalyst, Harvard University, and its affiliated academic healthcare centers, or the National Institutes of Health.
References 1 Adsay NV, Fukushima N, Furukawa T et al. Intraductal neoplasms of the pancreas. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, eds. WHO classification of tumours of the digestive system. 4th edn. Lyon: IARC; 2010: 304 – 313 2 Brugge WR, Lauwers GY, Sahani D et al. Cystic neoplasms of the pancreas. N Engl J Med 2004; 351: 1218 – 1226 3 Yoon WJ, Lee JK, Lee KH et al. Cystic neoplasms of the exocrine pancreas: an update of a nationwide survey in Korea. Pancreas 2008; 37: 254 – 258 4 Basturk O, Coban I, Adsay NV. Pancreatic cysts: pathologic classification, differential diagnosis, and clinical implications. Arch Pathol Lab Med 2009; 133: 423 – 438 5 Zhang XM, Mitchell DG, Dohke M et al. Pancreatic cysts: depiction on single-shot fast spin-echo MR images. Radiology 2002; 223: 547 – 553 6 Spinelli KS, Fromwiller TE, Daniel RA et al. Cystic pancreatic neoplasms: observe or operate. Ann Surg 2004; 239: 651 – 657; discussion 657659 7 Laffan TA, Horton KM, Klein AP et al. Prevalence of unsuspected pancreatic cysts on MDCT. AJR Am J Roentgenol 2008; 191: 802 – 807 8 Furukawa T, Kloppel G, Volkan Adsay N et al. Classification of types of intraductal papillary-mucinous neoplasm of the pancreas: a consensus study. Virchows Arch 2005; 447: 794 – 799 9 Furukawa T, Hatori T, Fujita I et al. Prognostic relevance of morphological types of intraductal papillary mucinous neoplasms of the pancreas. Gut 2011; 60: 509 – 516 10 Mino-Kenudson M, Fernandez-del Castillo C, Baba Y et al. Prognosis of invasive intraductal papillary mucinous neoplasm depends on histological and precursor epithelial subtypes. Gut 2011; 60: 1712 – 1720
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
1076
11 Distler M, Kersting S, Niedergethmann M et al. Pathohistological subtype predicts survival in patients with intraductal papillary mucinous neoplasm (IPMN) of the pancreas. Ann Surg 2013; 258: 324 – 330 12 Xiao HD, Yamaguchi H, Dias-Santagata D et al. Molecular characteristics and biological behaviours of the oncocytic and pancreatobiliary subtypes of intraductal papillary mucinous neoplasms. J Pathol 2011; 224: 508 – 516 13 Brugge WR, Lewandrowski K, Lee-Lewandrowski E et al. Diagnosis of pancreatic cystic neoplasms: a report of the cooperative pancreatic cyst study. Gastroenterology 2004; 126: 1330 – 1336 14 Pitman MB, Michaels PJ, Deshpande V et al. Cytological and cyst fluid analysis of small (< or = 3 cm) branch duct intraductal papillary mucinous neoplasms adds value to patient management decisions. Pancreatology 2008; 8: 277 – 284 15 Cizginer S, Turner B, Bilge AR et al. Cyst fluid carcinoembryonic antigen is an accurate diagnostic marker of pancreatic mucinous cysts. Pancreas 2011; 40: 1024 – 1028 16 Hara T, Ikebe D, Odaka A et al. Preoperative histological subtype classification of intraductal papillary mucinous neoplasms (IPMN) by pancreatic juice cytology with MUC stain. Ann Surg 2013; 257: 1103 – 1111 17 Yokoyama S, Kitamoto S, Higashi M et al. Diagnosis of pancreatic neoplasms using a novel method of DNA methylation analysis of mucin expression in pancreatic juice. PLoS One 2014; 9: e93760 18 Tanaka M, Chari S, Adsay V et al. International consensus guidelines for management of intraductal papillary mucinous neoplasms and mucinous cystic neoplasms of the pancreas. Pancreatology 2006; 6: 17 – 32 19 Pitman MB, Lewandrowski K, Shen J et al. Pancreatic cysts: preoperative diagnosis and clinical management. Cancer Cytopathol 2010; 118: 1 – 13 20 Pitman MB, Centeno BA, Daglilar ES et al. Cytological criteria of highgrade epithelial atypia in the cyst fluid of pancreatic intraductal papillary mucinous neoplasms. Cancer Cytopathol 2014; 122: 40 – 47 21 Pitman MB, Centeno BA, Genevay M et al. Grading epithelial atypia in endoscopic ultrasound-guided fine-needle aspiration of intraductal papillary mucinous neoplasms: an international interobserver concordance study. Cancer Cytopathol 2013; 121: 729 – 736
22 Tanaka M, Fernandez-del Castillo C, Adsay V et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012; 12: 183 – 197 23 Ban S, Naitoh Y, Mino-Kenudson M et al. Intraductal papillary mucinous neoplasm (IPMN) of the pancreas: its histopathologic difference between 2 major types. Am J Surg Pathol 2006; 30: 1561 – 1569 24 Hirono S, Tani M, Kawai M et al. The carcinoembryonic antigen level in pancreatic juice and mural nodule size are predictors of malignancy for branch duct type intraductal papillary mucinous neoplasms of the pancreas. Ann Surg 2012; 255: 517 – 522 25 Maire F, Voitot H, Aubert A et al. Intraductal papillary mucinous neoplasms of the pancreas: performance of pancreatic fluid analysis for positive diagnosis and the prediction of malignancy. Am J Gastroenterol 2008; 103: 2871 – 2877 26 Correa-Gallego C, Warshaw AL, Fernandez-del Castillo C. Fluid CEA in IPMNs: a useful test or the flip of a coin? Am J Gastroenterol 2009; 104: 796 – 797 27 Kucera S, Centeno BA, Springett G et al. Cyst fluid carcinoembryonic antigen level is not predictive of invasive cancer in patients with intraductal papillary mucinous neoplasm of the pancreas. JOP 2012; 13: 409 – 413 28 Duffy MJ. Carcinoembryonic antigen as a marker for colorectal cancer: is it clinically useful? Clin Chem 2001; 47: 624 – 630 29 Zhou J, Zhang L, Gu Y et al. Dynamic expression of CEACAM7 in precursor lesions of gastric carcinoma and its prognostic value in combination with CEA. World J Surg Oncol 2011; 9: 172 30 Hruban RH, Pitman MB, Klimstra D. Tumors of the pancreas. Atlas of tumor pathology. 4th series, fascicle 6. Washington, D.C.: Armed Forces Institute of Pathology; 2007 31 Ahnen DJ, Nakane PK, Brown WR. Ultrastructural localization of carcinoembryonic antigen in normal intestine and colon cancer: abnormal distribution of CEA on the surfaces of colon cancer cells. Cancer 1982; 49: 2077 – 2090 32 Hamada Y, Yamamura M, Hioki K et al. Immunohistochemical study of carcinoembryonic antigen in patients with colorectal cancer. Correlation with plasma carcinoembryonic antigen levels. Cancer 1985; 55: 136 – 141
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Original article
Copyright of Endoscopy is the property of Georg Thieme Verlag Stuttgart and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
1071
Authors
Won Jae Yoon1, 2, Ebubekir S. Daglilar1, Mari Mino-Kenudson3, Vicente Morales-Oyarvide3, Martha B. Pitman3, William R. Brugge4
Institutions
Institutions are listed at the end of article.
submitted 20. January 2014 accepted after revision 30. June 2014
Background and study aims: Intraductal papillary mucinous neoplasm (IPMN) consists of four epithelial subtypes. There are limited data on the endoscopic ultrasound (EUS) findings and/or cyst fluid analysis of the epithelial subtypes. The objective of this study was to determine whether there are differences in EUS and cyst fluid characteristics (carcinoembryonic antigen [CEA] concentration and cytology) among the subtypes. Patients and methods: The study cohort consisted of 85 patients (median age 68 years, 40 men) with resected and histologically confirmed branchduct or mixed-type IPMNs who underwent preoperative EUS-guided fine-needle aspiration between 1999 and 2010 for the evaluation of pancreatic cysts. EUS and cyst fluid characteristics were analyzed retrospectively and correlated with the subtypes. Results: The numbers of evaluated cystic lesions were 1 in 79 patients, 2 in 5 patients, and 3 in 1
patient. Of 92 IPMNs analyzed, gastric-type IPMNs were the most common (n = 68, 73.9 %), followed by intestinal (n = 17, 18.5 %), oncocytic (n = 5, 5.4 %), and pancreatobiliary subtypes (n = 2, 2.2 %). Gastric-type IPMNs were significantly smaller (cutoff 30 mm; P = 0.002), and less likely than other subtypes to have a mass lesion or mural nodule (P = 0.046) on EUS. Cyst fluid CEA concentration varied among the subtypes (median concentrations for gastric, intestinal, oncocytic, and pancreatobiliary types 619.8, 83.0, 5.1, and 270.0 ng/mL, respectively; P = 0.012). The presence of neoplastic epithelial cells (P = 0.624) and extracellular mucin (P = 0.208) on cytology had no association with subtypes. Conclusions: Gastric-type IPMNs, the most common subtype, are characterized by high concentrations of cyst fluid CEA, small cyst diameter, and low risk EUS imaging features.
Introduction
teins, IPMNs are classified into four epithelial subtypes: gastric, intestinal, oncocytic, and pancreatobiliary [8]. The epithelial subtypes are reported to be significantly associated with a variety of clinicopathological features such as sex, age, histological grade, differential involvement of the pancreatic duct system, invasive phenotypes, stage, recurrence, and patient survival [9 – 12]. Gastric-type IPMNs are reported to have a more favorable prognosis compared with other epithelial subtypes [9]. Endoscopic ultrasound (EUS) is being increasingly utilized in the diagnosis of IPMN. EUS provides detailed morphological analysis of PCLs. EUSguided fine-needle aspiration (EUS-FNA) enables cyst fluid analysis such as carcinoembryonic antigen (CEA) and cytology [13, 14]. Cyst fluid CEA concentration has been reported to be the most accurate marker to differentiate mucinous (i. e. IPMN and mucinous cystic neoplasm) from nonmucinous PCLs [13, 15].
Bibliography DOI http://dx.doi.org/ 10.1055/s-0034-1377629 Published online: 10.9.2014 Endoscopy 2014; 46: 1071– 1077 © Georg Thieme Verlag KG Stuttgart · New York ISSN 0013-726X Corresponding author William R. Brugge, MD Gastrointestinal Unit Harvard Medical School Massachusetts General Hospital 55 Fruit Street Boston, MA 02114 United States Fax: +1-617-724-5997 [email protected]
!
Intraductal papillary mucinous neoplasm of the pancreas (IPMN) is defined as a grossly visible, intraductal, epithelial neoplasm composed of mucin-producing cells, arising in the main pancreatic duct and/or its branches [1]. The incidence of IPMN is reported to be approximately 5 % among pancreatic resection specimens and 20 % – 40 % among all pancreatic cystic neoplasms [1 – 4]. However, these figures represent surgical data, and therefore the true incidence of IPMNs is most likely to be higher. Indeed, the prevalence of pancreatic cystic lesions (PCLs) ranges between 1.2 % and 19.6 % according to image-based studies [5 – 7], and a significant proportion of these lesions are probably IPMNs. Recently, morphological variations in the epithelial cell types of IPMNs have been recognized. Based on histomorphological features and immunohistochemical reactivity for mucin glycopro-
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Characterization of epithelial subtypes of intraductal papillary mucinous neoplasm of the pancreas with endoscopic ultrasound and cyst fluid analysis
Original article
Determination of the epithelial subtypes of IPMN is only possible through the histological evaluation of the resected specimens. The use of pancreatic juice obtained during endoscopic retrograde cholangiopancreatography for cytology or DNA methylation analysis of mucin expression for the prediction of epithelial subtypes of IPMN has been reported [16, 17]. However, to the best of our knowledge, there are no available data regarding EUS findings and/or cyst fluid characteristics of the epithelial subtypes. Determination of such association may aid in the prediction of the clinicopathological behavior of the lesion, and furthermore, in the planning of treatment. The aims of this study were to determine 1) EUS imaging features, 2) cyst fluid CEA concentration, and 3) the cytological characteristics of the epithelial subtypes of IPMN. It was hypothesized that such characteristics might be associated with particular epithelial subtypes.
Patients and methods !
Study population The study was approved by the Partners Human Research Committee Institutional Review Board. A prospective database was retrospectively analyzed to identify all patients who underwent EUS-FNA to evaluate PCLs at Massachusetts General Hospital. The study cohort consisted of patients who underwent surgery and were ultimately diagnosed with IPMNs by histopathological analysis between 1999 and 2010. The decision to undergo a surgical resection was made based on the 2006 international consensus guideline in the majority of cases [18]. These parameters (presumed main-duct type or mixed-type IPMN, branch-duct IPMN [BD-IPMN] > 3 cm in diameter, presence of a mass lesion or mural nodule, positive cytology, and/or presence of symptoms) had been followed for a few years at Massachusetts General Hospital before the guidelines were published. Other patients were referred for resection based on the recommendation of their physician or surgeon. The following parameters were analyzed: sex, age, EUS and FNA characteristics (location, size in long diameter, presence of septation, mass lesion, mural nodule, and thick cyst wall, and cyst fluid viscosity as reported by endosonographer), cyst fluid CEA concentration, cyst fluid cytology results (presence of neoplastic epithelial cells or extracellular mucin, and cytological grade of dysplasia), and surgical histology results (epithelial subtype, type of duct involvement, and grade of dysplasia). For patients with multiple PCLs evaluated with EUS-FNA, each lesion was independently analyzed. In such patients, the location and the morphological features including the size of the cyst were carefully compared between the EUS findings and surgical specimens. Patients were excluded from the study if the time interval between EUS-FNA and surgery was more than 1 year, and in cases in which EUS showed distinctive cystic lesions but histology confirmed main-duct type IPMN (MD-IPMN). These criteria were applied so that the results of EUS-FNA would reflect the cyst fluid characteristics in a timely manner, and to exclude the non-neoplastic branch-duct dilation secondary to MD-IPMNs.
EUS-FNA procedure All EUS-FNAs were performed as previously described [13, 15]. In brief, EUS was performed using a curvilinear echoendoscope by three experienced endosonographers. The PCL was aspirated under EUS guidance. Vascular structures were identified and avoided using Doppler imaging. PCLs in the pancreatic head or
uncinate process were accessed via a transduodenal approach; those in the body or tail were aspirated via a transgastric approach. To avoid infection of the PCL, intravenous antibiotics were administered during the procedure; an oral quinolone was administered for 2 – 3 days after the procedure in selected cases. When adequate amount of cyst fluid was aspirated, the sample was sent for both CEA concentration and cytology. When limited amount of cyst fluid was aspirated, the sample was triaged according to the clinical question. If the primary question was whether the PCL was mucinous or nonmucinous, the priority was given to CEA concentration. If the primary question was whether the PCL was malignant or benign, the priority was given to cytology [19].
Cyst fluid CEA concentration Cyst fluid CEA concentration was measured as previously described [13, 14]. To summarize, undiluted cyst fluid obtained from EUS-FNA was analyzed for CEA by a commercially available radioimmunoassay kit (Abbott Laboratory, Diagnostics Division, Abbott Park, Illinois, USA). When CEA concentration of a PCL was reported as a range, the lowest value of the range was considered to be the CEA level. For example, when cyst fluid CEA concentration was reported to be > 500 ng/mL, the cyst fluid CEA concentration was considered to be 500 ng/mL.
Cytological examination of the cyst fluid Cytology specimens consisted of direct smears, cytospins or liquid-based preparations. They were routinely stained with Papanicolaou stain. Rarely, hematoxylin and eosin stains were used for rapid interpretation. Mucicarmine and alcian blue stains were used to identify mucin in cyst aspirates with an adequate amount of fluid. In the cytology reports of 24 PCLs, the results of mucin stains were not available; these lesions were considered to have no extracellular mucin present. Epithelial cells in the cyst fluid were evaluated for cytological atypia. In brief, the cytology slides were re-reviewed by a cytopathologist and the epithelial cells were classified in a two-tiered system of low grade or high grade epithelial atypia. Low grade epithelial atypia represented cells with low to intermediate grade dysplasia or cells indeterminate for low grade dysplasia vs. gastrointestinal contamination. High grade epithelial atypia represented cells with either high grade dysplasia or invasive carcinoma [20, 21].
Histological examination of IPMN The pathology slides from the surgical specimen were reviewed by a single pathologist. Epithelial subtype, grade of dysplasia, and type of duct involvement by IPMN were evaluated. In patients with multiple cysts resected, all the specimens were reviewed. The epithelial subtype was determined on the basis of histomorphology and mucin glycoprotein immunostaining [8]. The grade of dysplasia was reported based on the highest degree of the dysplasia identified. The type of duct involvement was determined based on the results of pathology and imaging studies [10]. If PCLs had a pure single epithelial subtype, it was recorded as such. In cases exhibiting heterogeneous epithelium, the predominant type of the highest grade was recorded as the subtype and used for analysis [10]. Only invasive carcinoma was considered to be malignant, as stated in both the World Health Organization classification and the 2012 international consensus guideline [1, 22].
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
1072
Original article
Histological characteristics of the PCLs
Patient characteristics.
Number of patients
85
Sex, male/female, n
40/45
Age, median (range), years
68 (39 – 83)
Cystic lesions evaluated, n
92
Patients with lesions, n (%) 1 cystic lesion
79 (92.9)
2 cystic lesions
5 (5.9)
3 cystic lesions
1 (1.2) 53 (4 – 246)
Time between EUS-FNA and surgery, median (range), days EUS-FNA, endoscopic ultrasound-guided fine-needle aspiration.
Statistical analysis All continuous variables are reported as the median (range). For categorical data, the chi-squared test or the Fisher’s exact test was performed, where appropriate. For comparison of cyst fluid CEA concentration, the Kruskal – Wallis test was performed. A two-sided P value of < 0.05 was considered to be statistically significant. STATA 12.1 (StataCorp LP, College Station, Texas, USA) was used for all statistical analyses.
Of the 85 patients, 62 had gastric, 17 had intestinal, 4 had oncocytic, and 2 had the pancreatobiliary subtype IPMN. A PCL with mixed epithelial subtype was identified in four patients. The mixed subtypes were pancreatobiliary-oncocytic (n = 1), pancreatobiliary-gastric (n = 1), gastric-intestinal (n = 1), and intestinal-gastric (n = 1). They were considered to be pancreatobiliary, pancreatobiliary, gastric, and intestinal subtypes, respectively, based on the grade and predominance of the subtypes within the tumor. BD-IPMNs were diagnosed in 40 gastric, 8 intestinal, 2 oncocytic, and no pancreatobiliary IPMN patients; the other patients (22 gastric, 9 intestinal, 2 oncocytic, and 2 pancreatobiliary IPMN patients) had mixed-type IPMNs. In all but one patient with multiple PCLs aspirated, the cysts were of the same grade of dysplasia and epithelial subtype. In one patient who had two cysts aspirated, pathology revealed low grade dysplasia in one cyst and IPMN with an associated invasive carcinoma in the other. Epithelial subtypes were significantly associated with the grade of dysplasia (P < 0.001). Subsequent comparison showed that gastric-type IPMNs were significantly associated with noninvasive histology compared with other types of IPMNs (92.6 % for gastric-type vs. 66.7 % for nongastric-type IPMNs; P = 0.004). The " Table 2. histological characteristics are summarized in ●
EUS features Results !
Patient characteristics A total of 1060 patients underwent EUS-FNA for the evaluation of PCLs during the study period. A total of 328 patients underwent resection of IPMNs during the same period. A total of 93 patients with IPMNs who underwent preoperative EUS-FNA of distinctive cystic lesions were identified. Eight patients were excluded from the analysis: five had a time interval between EUS-FNA and surgery of more than 1 year, and three had MD-IPMN on histology. Thus, a total of 85 patients (40 male) with 92 distinctive PCLs were evaluated. The median age at diagnosis was 68 years. The numbers of evaluated PCLs were 1 in 79 patients, 2 in 5 patients, and 3 in 1 patient. The median time interval between EUS-FNA " Table 1). and surgery was 53 days (●
Table 2
The epithelial subtype was associated with the presence of a mass lesion or mural nodule (P = 0.012) and size of the PCL (P = 0.003). Subsequent comparison showed that gastric-type IPMNs were significantly less likely to demonstrate a mass lesion or mural nodule on EUS than other subtypes of IPMNs (17.6 % for gastric-type vs. 37.5 % for nongastric type; P = 0.046). In addition, gastric-type IPMNs were less likely to be ≥ 30 mm in size than other subtypes of IPMNs (21.5 % for gastric-type vs. 57.1 % for nongastric type; P = 0.002). However, the epithelial subtype was not associated with the location of the PCL, presence of a thick cyst wall, septation, or viscous fluid on FNA " Table 3). (●
Histological characteristics of intraductal papillary mucinous neoplasm.
Epithelial subtype
Total
Gastric
Intestinal
Oncocytic
Pancreatobiliary
Patients, n (%)
62 (72.9)
17 (20.0)
4 (4.7)
2 (2.4)
85 (100.0)
Cystic lesions, n (%)
68 (73.9)
17 (18.5)
5 (5.4)
2 (2.2)
92 (100.0)
Branch duct
46 (67.6)
8 (47.1)
2 (40.0)
0 (0)
56 (60.9)
Mixed type
22 (32.4)
9 (52.9)
3 (60.0)
2 (100.0)
36 (39.1)
Low
26 (38.2)
0 (0)
0 (0)
0 (0)
26 (28.3)
Intermediate
39 (42.4)
Macroscopic type, n (%) 2
P value1
0.066
Grade of dysplasia, n (%) 2,3
< 0.001 32 (47.0)
7 (41.2)
0 (0)
0 (0)
High
5 (7.4)
6 (35.3)
2 (40.0)
1 (50.0)
14 (15.2)
Invasive
5 (7.4)
4 (23.5)
3 (60.0)
1 (50.0)
13 (14.1)
IPMN, intraductal papillary mucinous neoplasm. 1 Freeman–Halton extension of the Fisher’s exact test of the four epithelial subtypes. 2 Number and percentage of cystic lesions. 3 Gastric-type IPMNs were significantly associated with noninvasive histology compared with other subtypes of IPMNs (92.6 % gastric-type vs. 66.7 % for nongastric-type IPMNs; P = 0.004, Fisher’s exact test).
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Table 1
1073
Original article
Endoscopic ultrasound characteristics of cystic lesions according to epithelial subtype of intraductal papillary mucinous neoplasm.
Epithelial subtype Gastric Mass lesion/mural nodule, n (%)
P value1
Total
Intestinal
Oncocytic
Pancreatobiliary
2
0.012
Yes
12 (17.6)
4 (23.5)
4 (80.0)
1 (50.0)
21 (22.8)
No
56 (82.4)
13 (76.5)
1 (20.0)
1 (50.0)
71 (77.2)
≥ 30 mm
14 (21.5)
8 (50.0)
3 (100.0)
1 (50.0)
26 (30.2)
< 30 mm
51 (78.5)
8 (50.0)
0 (0.0)
1 (50.0)
60 (69.8)
Head
39 (57.4)
11 (64.7)
3 (60.0)
1 (50.0)
54 (58.7)
Body-tail
29 (42.6)
6 (35.3)
2 (40.0)
1 (50.0)
38 (41.3)
Yes
11 (16.2)
4 (23.5)
2 (40.0)
0 (0.0)
17 (18.5)
No
57 (83.8)
13 (76.5)
3 (60.0)
2 (100.0)
75 (81.5)
Yes
38 (55.9)
9 (52.9)
4 (80.0)
1 (50.0)
52 (56.5)
No
30 (44.1)
8 (47.1)
1 (20.0)
1 (50.0)
40 (43.5)
Yes
38 (82.6)
14 (100.0)
2 (100.0)
2 (100.0)
56 (87.5)
No
8 (17.4)
0 (0.0)
0 (0.0)
Size, n (%) 3,4
0.003
Location, n (%)
0.962
Thick cyst wall, n (%)
0.395
Septation, n (%)
0.796
Viscous fluid, n (%) 5
0.407 0 (0.0)
8 (12.5)
IPMN, intraductal papillary mucinous neoplasm. 1 Freeman–Halton extension of the Fisher’s exact test of the four epithelial subtypes. 2 Gastric-type IPMNs were significantly less likely to demonstrate a mass lesion or mural nodule on EUS than other subtypes of IPMNs (P = 0.046, chi-squared test). 3 Results of 86 cystic lesions with size reported. 4 Gastric-type IPMNs were less likely to be ≥ 30 mm in size than other subtypes of IPMNs (P = 0.002, chi-squared test). 5 Results of 64 cystic lesions with fluid viscosity reported.
Cyst fluid CEA concentration Cyst fluid CEA concentrations were reported in 60 PCLs (49 gastric, 6 intestinal, 4 oncocytic, and 1 pancreatobiliary). The epithelial subtype was significantly associated with cyst fluid CEA concentration (P = 0.012, Kruskal–Wallis test), with the gastric-type IPMN apparently having the highest median concentration " Fig. 1, ● " Table 4). However, the cyst fluid CEA concentration (● was not significantly associated with the grade of dysplasia of " Table 4). IPMNs (P = 0.072) (●
Cyst fluid cytology Cyst fluid cytology was performed in 76 PCLs. Neoplastic epithelial cells were identified in 28 lesions (36.8 %). High grade epithelial atypia was present in 21 of 28 lesions. Extracellular mucin was present in 42 lesions (55.3 %). When the presence of mucin was defined as viscous fluid as described by endosonographer or presence of extracellular mucin on cytological specimen, mucin was present in 72 lesions (78.3 %). The presence of neoplastic epithelial cells or extracellular mucin was not associated with epithelial subtype. The high grade epithelial atypia features showed a tendency to be associated with epithelial subtype (P = " Table 5. 0.092). The cytology results are summarized in ●
Discussion !
The purpose of this study was to analyze the EUS-FNA findings of the epithelial subtypes of IPMN. Analysis of 92 PCLs showed that the gastric-type IPMN was positively associated with high concentrations of cyst fluid CEA and negatively associated with high risk EUS imaging features of mass lesion or mural nodule and large cyst size. Cytological features of cells exfoliated into the cyst fluid were nonspecific to the epithelial subtype.
100000
10000
1000
100
10
1 Gastric
Intestinal
Oncocytic
Pancreatobiliary
Fig. 1 Cyst fluid carcinoembryonic antigen (CEA) concentrations (log scale) according to epithelial subtypes of intraductal papillary mucinous neoplasms (n = 60). Epithelial subtype was significantly associated with cyst fluid CEA concentration (P = 0.012, Kruskal–Wallis test).
IPMNs have been known to have morphological variations that are associated with clinicopathological outcomes [23]. A consensus classification for distinguishing the four epithelial subtypes of IPMN was drawn in 2005, and is based on the morphological features and the immunohistochemical reactivity for mucin glycoproteins [8]. This classification was validated in recent studies [9, 10]. In a large-scale study of 283 cases, the epithelial subtypes were associated with sex, age, duct dilation, detection of mural nodule on imaging, histological grade, pancreatic duct involvement, invasive phenotype, stage, recurrence, and patient survival [9]. Moreover, Mino-Kenudson et al. reported that the prognosis
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Table 3
CEA concentration (ng/mL)
1074
Original article
Table 6 Characteristics of epithelial subtypes of intraductal papillary mucinous neoplasm.
Table 4 Comparison of cyst fluid carcinoembryonic antigen concentrations according to epithelial subtype and grade of dysplasia of intraductal papillary mucinous neoplasm.
Epithelial subtype
0.012 619.8 (22.3 – 50000.0)
Intestinal (n = 6)
83.0 (11.1 – 5165.0)
Oncocytic (n = 4)
5.1 (3.4 – 358.8)
Pancreatobiliary (n = 1)
270.0
Low (n = 23)
619.8 (26.7 – 50000.0)
Intermediate (n = 24)
343.4 (11.1 – 19704.0)
High (n = 6)
185.8 (5.6 – 1872.0)
Invasive (n = 7)
98.4 (3.4 – 5165.0)
Mucin glyco-
Correlated
protein staining
invasive IPMN
1
0.072 1
Grade of dysplasia
Architecture
subtype
Gastric
Thick finger-like papillae
MUC5AC
Tubular adenocarcinoma
Intestinal
Villous papillae
MUC2, MUC5AC
Colloid carcinoma
Oncocytic
Thick branching complex papillae
MUC1, MUC5AC
Oncocytic carcinoma
Pancreatobiliary
Thin branching complex papillae
MUC1, MUC5AC
Tubular adenocarcinoma
IPMN, intraductal papillary mucinous neoplasm.
CEA, carcinoembryonic antigen. 1 Kruskal–Wallis test.
of invasive IPMN was shown to be dependent upon the histology of invasive components and precursor epithelial subtypes [10]. In addition, the histology of invasive components of IPMN was closely associated with the precursor epithelial subtypes. Colloid and oncocytic carcinomas originated only from intestinal and oncocytic IPMN types, respectively, and were frequently of the main-duct type; tubular adenocarcinomas primarily arose from the gastric-type IPMN, and were often associated with BD-IPMN. The colloid and oncocytic types of invasive IPMNs were more indolent compared with tubular types. The authors concluded that epithelial subtypes of precursor IPMN correlate with the histological subtype of invasive carcinoma, and ultimately with prognosis [10]. It is generally accepted that the gastric-type IPMN has a favorable outcome. However, when a carcinoma does develop in gastric-type IPMN, it is usually of the tubular type and demonstrates an aggressive behavior [22]. A brief summary of the fea" Table 6. tures of epithelial subtypes is presented in ● In the current study, gastric-type IPMNs were less likely to exhibit a mass lesion or mural nodule on EUS, and were smaller than other subtypes of IPMNs. This is analogous to the association between the epithelial subtype and image findings reported by Furukawa et al., where the intestinal-type IPMNs were larger than gastric-type IPMNs, and oncocytic and pancreatobiliary subtypes were associated with detection of mural nodules [9].
Table 5
Interestingly, cyst fluid CEA concentration varied significantly with the epithelial subtypes of IPMN. However, there was no statistically significant association between the cyst fluid CEA concentration and the grade of dysplasia. For IPMNs, the ability of cyst fluid CEA to differentiate malignant from benign PCLs has been controversial. Although a recent report indicated that a high CEA concentration in the pancreatic juice is associated with malignancy in BD-IPMNs [24], these results are difficult to be applied directly to the results of EUSFNA, as there are issues of differences in the specimen (cyst fluid vs. pancreatic juice) and the sampling method (EUS-FNA vs. pancreatic juice aspiration during endoscopic retrograde pancreatography). A study of 41 IPMN patients who underwent preoperative EUS-FNA reported that a cyst fluid CEA concentration > 200 ng/mL had sensitivity, specificity, positive predictive value, and negative predictive value of 90 %, 71 %, 50 %, and 96 %, respectively, for the diagnosis of malignant IPMN [25]. However, these findings were not reproduced in subsequent studies. In a report of 72 IPMN patients, the cutoff cyst fluid CEA concentration of 200 ng/mL demonstrated sensitivity, specificity, positive predictive value, and negative predictive value of 47 %, 40 %, 20 %, and 70 %, respectively [26]. Another study by Kucera et al. involving 47 IPMN patients demonstrated that the cyst fluid CEA concentration had no association with grades of dysplasia [27]. Results
Comparison of cytology characteristics according to epithelial subtype of intraductal papillary mucinous neoplasm. Results of 76 cystic lesions.
Epithelial subtype
Cystic lesions, n
Total
Gastric
Intestinal
Oncocytic
Pancreatobiliary
54
15
5
2
76
Neoplastic epithelial cells, n (%)
P value1
0.624
Yes
22 (40.7)
4 (26.7)
1 (20.0)
1 (50.0)
28 (36.8)
No
32 (59.3)
11 (73.3)
4 (80.0)
1 (50.0)
48 (63.2)
High grade epithelial atypia 3
18 (81.8)
1 (25.0)
1 (100.0)
1 (100.0)
21 (75.0)
Low grade epithelial atypia 4
4 (18.2)
3 (75.0)
0 (0.0)
0 (0.0)
7 (25.0)
Yes
28 (51.9)
11 (73.3)
3 (60.0)
0 (0.0)
42 (55.3)
No
26 (48.1)
4 (26.7)
2 (40.0)
2 (100.0)
34 (44.7)
Cytological atypia, n (%) 2
0.092
Extracellular mucin, n (%)
0.208
1
Freeman–Halton extension of the Fisher’s exact test of the four epithelial subtypes. Results from 28 cystic lesions with neoplastic epithelial cells present in the cyst fluid. High grade epithelial atypia represents cells with either high grade dysplasia or invasive carcinoma. 4 Low grade epithelial atypia represents cells with low to intermediate grade dysplasia or cells indeterminate for low grade dysplasia vs. gastrointestinal contamination. 2 3
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Gastric (n = 49)
Epithelial
P value
CEA, median (range), ng/mL
1075
Original article
from the current study suggest that the cyst fluid CEA concentration may be correlated with epithelial subtypes rather than grade of dysplasia. The mechanisms of the difference in the cyst fluid CEA according to the epithelial subtype of IPMN remain to be elucidated. One hypothesis is that the quantitative expression of CEA would be different according to epithelial subtypes. Although CEA is most well known to be expressed in colorectal cancer tissues that share the immunophenotype with intestinal-type IPMNs [28], it is also reported to be expressed in gastric cancer and precursor lesions of gastric cancer [29], and pancreatic ductal adenocarcinoma [30]. Thus, each epithelial subtype of IPMN might have the potential to express CEA. There might be a difference in the quantity of CEA expression according to epithelial subtypes, which may result in different cyst fluid CEA concentrations. Another hypothesis would be the difference in the polarized distribution of CEA according to epithelial subtypes. In colorectal cancer, CEA is reported to be distributed along the luminal border in well-differentiated cancers, whereas it is distributed over the entire cell surface in a diffuse pattern in poorly differentiated cancers [31]. The distribution of CEA is correlated with the plasma CEA level in colorectal cancer. Basolateral distribution of CEA was associated with high plasma CEA level [32]. In IPMNs, luminal distribution of CEA might result in high cyst fluid CEA concentration, whereas basolateral distribution of CEA might result in low concentration of cyst fluid CEA. The presence of neoplastic epithelial cells and extracellular mucin on cytology of cyst fluid was not associated with epithelial subtypes of IPMN. Cytological examination of the cyst fluid mainly evaluates the neoplastic epithelial cells shed into the cyst. Thus, one may expect that malignant lesions would shed more neoplastic epithelial cells than benign ones [19]. Given the fact that gastric-type IPMNs usually exhibit low to intermediate grade dysplasia, and gastric-type IPMNs may have an abundant flat cyst lining, as opposed to other subtypes which predominantly demonstrate papillary architecture [8], it was expected that the presence of neoplastic epithelial cells in cyst fluid would be more frequent in nongastric-type IPMNs. However, no such association was found. Contrary to expectations, although statistically insignificant, there was a trend for gastric-type IPMNs to have neoplastic cells in the FNA specimens more frequently than other epithelial subtypes. One suggested hypothesis to explain this observation is that the epithelium of gastric-type IPMN might shed more neoplastic cells into the cyst fluid. Another hypothesis is that the detached gastric-type neoplastic cells might maintain morphological integrity longer than nongastric subtypes. This study has limitations, including the retrospective and singlecenter nature of the study. In addition, only the lesions that had been surgically resected could be included in the study, resulting in an inevitable selection bias. Reflecting the proportion of the epithelial subtypes of IPMN, the numbers of patients with oncocytic and pancreatobiliary subtypes were small. The proportion of gastric-type IPMN was somewhat higher than in other studies. This may be partially explained by the high proportion of BDIPMNs in the study cohort. One might argue that the MD-IPMN cases should have been included in the analysis. However, the results of Kucera et al. [27] and Hirono et al. [24] suggest that the specimen for fluid analysis should be aspirated from the primary PCL with neoplastic epithelial lining. The features of non-neoplastic branch duct dilation secondary to MD-IPMNs may not be representative of the main neoplasm. To the best of our knowl-
edge, this is the first study to determine the EUS and cyst fluid findings of the epithelial subtypes of IPMN. In conclusion, the cyst fluid CEA concentration seems to be associated with the epithelial subtypes of IPMN; it seems to have a tendency toward negative correlation with the grade of dysplasia. Gastric-type IPMNs, which are strongly associated with low histological grade and are the most common type, are characterized by high concentrations of cyst fluid CEA, small cyst diameter, and low risk EUS imaging features. These features may aid in the prediction of the epithelial subtypes of IPMN and provide valuable information for selection of treatment strategy. Competing interests: Professor Brugge has received research funding from RedPath and Asuragen. Institutions 1 Gastrointestinal Unit, Massachusetts General Hospital, Boston, Massachusetts, United States 2 Department of Internal Medicine, Inje University Seoul Paik Hospital, Inje University College of Medicine, Seoul, Republic of Korea 3 Department of Pathology, Massachusetts General Hospital, Boston, Massachusetts, United States 4 Gastrointestinal Unit, Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts, United States
Acknowledgments !
This work was conducted with support from Harvard Catalyst, The Harvard Clinical and Translational Science Center (National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health Award 8UL1TR000170-05 and financial contributions from Harvard University and its affiliated academic healthcare centers). The content is solely the responsibility of the authors and does not necessarily represent the official views of Harvard Catalyst, Harvard University, and its affiliated academic healthcare centers, or the National Institutes of Health.
References 1 Adsay NV, Fukushima N, Furukawa T et al. Intraductal neoplasms of the pancreas. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, eds. WHO classification of tumours of the digestive system. 4th edn. Lyon: IARC; 2010: 304 – 313 2 Brugge WR, Lauwers GY, Sahani D et al. Cystic neoplasms of the pancreas. N Engl J Med 2004; 351: 1218 – 1226 3 Yoon WJ, Lee JK, Lee KH et al. Cystic neoplasms of the exocrine pancreas: an update of a nationwide survey in Korea. Pancreas 2008; 37: 254 – 258 4 Basturk O, Coban I, Adsay NV. Pancreatic cysts: pathologic classification, differential diagnosis, and clinical implications. Arch Pathol Lab Med 2009; 133: 423 – 438 5 Zhang XM, Mitchell DG, Dohke M et al. Pancreatic cysts: depiction on single-shot fast spin-echo MR images. Radiology 2002; 223: 547 – 553 6 Spinelli KS, Fromwiller TE, Daniel RA et al. Cystic pancreatic neoplasms: observe or operate. Ann Surg 2004; 239: 651 – 657; discussion 657659 7 Laffan TA, Horton KM, Klein AP et al. Prevalence of unsuspected pancreatic cysts on MDCT. AJR Am J Roentgenol 2008; 191: 802 – 807 8 Furukawa T, Kloppel G, Volkan Adsay N et al. Classification of types of intraductal papillary-mucinous neoplasm of the pancreas: a consensus study. Virchows Arch 2005; 447: 794 – 799 9 Furukawa T, Hatori T, Fujita I et al. Prognostic relevance of morphological types of intraductal papillary mucinous neoplasms of the pancreas. Gut 2011; 60: 509 – 516 10 Mino-Kenudson M, Fernandez-del Castillo C, Baba Y et al. Prognosis of invasive intraductal papillary mucinous neoplasm depends on histological and precursor epithelial subtypes. Gut 2011; 60: 1712 – 1720
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
1076
11 Distler M, Kersting S, Niedergethmann M et al. Pathohistological subtype predicts survival in patients with intraductal papillary mucinous neoplasm (IPMN) of the pancreas. Ann Surg 2013; 258: 324 – 330 12 Xiao HD, Yamaguchi H, Dias-Santagata D et al. Molecular characteristics and biological behaviours of the oncocytic and pancreatobiliary subtypes of intraductal papillary mucinous neoplasms. J Pathol 2011; 224: 508 – 516 13 Brugge WR, Lewandrowski K, Lee-Lewandrowski E et al. Diagnosis of pancreatic cystic neoplasms: a report of the cooperative pancreatic cyst study. Gastroenterology 2004; 126: 1330 – 1336 14 Pitman MB, Michaels PJ, Deshpande V et al. Cytological and cyst fluid analysis of small (< or = 3 cm) branch duct intraductal papillary mucinous neoplasms adds value to patient management decisions. Pancreatology 2008; 8: 277 – 284 15 Cizginer S, Turner B, Bilge AR et al. Cyst fluid carcinoembryonic antigen is an accurate diagnostic marker of pancreatic mucinous cysts. Pancreas 2011; 40: 1024 – 1028 16 Hara T, Ikebe D, Odaka A et al. Preoperative histological subtype classification of intraductal papillary mucinous neoplasms (IPMN) by pancreatic juice cytology with MUC stain. Ann Surg 2013; 257: 1103 – 1111 17 Yokoyama S, Kitamoto S, Higashi M et al. Diagnosis of pancreatic neoplasms using a novel method of DNA methylation analysis of mucin expression in pancreatic juice. PLoS One 2014; 9: e93760 18 Tanaka M, Chari S, Adsay V et al. International consensus guidelines for management of intraductal papillary mucinous neoplasms and mucinous cystic neoplasms of the pancreas. Pancreatology 2006; 6: 17 – 32 19 Pitman MB, Lewandrowski K, Shen J et al. Pancreatic cysts: preoperative diagnosis and clinical management. Cancer Cytopathol 2010; 118: 1 – 13 20 Pitman MB, Centeno BA, Daglilar ES et al. Cytological criteria of highgrade epithelial atypia in the cyst fluid of pancreatic intraductal papillary mucinous neoplasms. Cancer Cytopathol 2014; 122: 40 – 47 21 Pitman MB, Centeno BA, Genevay M et al. Grading epithelial atypia in endoscopic ultrasound-guided fine-needle aspiration of intraductal papillary mucinous neoplasms: an international interobserver concordance study. Cancer Cytopathol 2013; 121: 729 – 736
22 Tanaka M, Fernandez-del Castillo C, Adsay V et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012; 12: 183 – 197 23 Ban S, Naitoh Y, Mino-Kenudson M et al. Intraductal papillary mucinous neoplasm (IPMN) of the pancreas: its histopathologic difference between 2 major types. Am J Surg Pathol 2006; 30: 1561 – 1569 24 Hirono S, Tani M, Kawai M et al. The carcinoembryonic antigen level in pancreatic juice and mural nodule size are predictors of malignancy for branch duct type intraductal papillary mucinous neoplasms of the pancreas. Ann Surg 2012; 255: 517 – 522 25 Maire F, Voitot H, Aubert A et al. Intraductal papillary mucinous neoplasms of the pancreas: performance of pancreatic fluid analysis for positive diagnosis and the prediction of malignancy. Am J Gastroenterol 2008; 103: 2871 – 2877 26 Correa-Gallego C, Warshaw AL, Fernandez-del Castillo C. Fluid CEA in IPMNs: a useful test or the flip of a coin? Am J Gastroenterol 2009; 104: 796 – 797 27 Kucera S, Centeno BA, Springett G et al. Cyst fluid carcinoembryonic antigen level is not predictive of invasive cancer in patients with intraductal papillary mucinous neoplasm of the pancreas. JOP 2012; 13: 409 – 413 28 Duffy MJ. Carcinoembryonic antigen as a marker for colorectal cancer: is it clinically useful? Clin Chem 2001; 47: 624 – 630 29 Zhou J, Zhang L, Gu Y et al. Dynamic expression of CEACAM7 in precursor lesions of gastric carcinoma and its prognostic value in combination with CEA. World J Surg Oncol 2011; 9: 172 30 Hruban RH, Pitman MB, Klimstra D. Tumors of the pancreas. Atlas of tumor pathology. 4th series, fascicle 6. Washington, D.C.: Armed Forces Institute of Pathology; 2007 31 Ahnen DJ, Nakane PK, Brown WR. Ultrastructural localization of carcinoembryonic antigen in normal intestine and colon cancer: abnormal distribution of CEA on the surfaces of colon cancer cells. Cancer 1982; 49: 2077 – 2090 32 Hamada Y, Yamamura M, Hioki K et al. Immunohistochemical study of carcinoembryonic antigen in patients with colorectal cancer. Correlation with plasma carcinoembryonic antigen levels. Cancer 1985; 55: 136 – 141
Yoon Won Jae et al. IPMN subtypes: EUS and cyst fluid analysis … Endoscopy 2014; 46: 1071–1077
1077
This document was downloaded for personal use only. Unauthorized distribution is strictly prohibited.
Original article
Copyright of Endoscopy is the property of Georg Thieme Verlag Stuttgart and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
![[Intraductal papillary mucinous neoplasm of the pancreas].](/assets/img/hold.png)
