- Email: [email protected]
Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series
Pancreatology xxx (2017) 1e7
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series James F. Griffin, MD a, Andrew J. Page, MD a, Georges J. Samaha, MD a, Adrienne Christopher, BA a, Feriyl Bhaijee, MD b, Maryam K. Pezhouh, MD b, Niek A. Peters, BSc a, Ralph H. Hruban, MD b, Jin He, MD a, Martin A. Makary, MD a, Anne Marie Lennon, MD, PhD c, John L. Cameron, MD a, Christopher L. Wolfgang, MD, PhD a, Matthew J. Weiss, MD a, * a b c
Departments of Surgery, The Johns Hopkins University School of Medicine, Baltimore, MD, USA Departments of Pathology, The Johns Hopkins University School of Medicine, Baltimore, MD, USA Departments of Gastroenterology, The Johns Hopkins University School of Medicine, Baltimore, MD, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 November 2016 Received in revised form 16 March 2017 Accepted 7 April 2017 Available online xxx
Background/Objectives: Mucinous cystic neoplasms (MCNs) are rare pancreas tumors distinguished from intraductal papillary mucinous neoplasms (IPMNs) by the presence of ovarian-type stroma. Historical outcomes for MCNs vary due to previously ambiguous diagnostic criteria resulting in confusion with IPMNs. This study seeks to characterize and clarify the clinical features and long-term outcomes of MCNs versus IPMNs in the largest single-institution series of pathology-confirmed MCNs to date. Methods: We compared 142 MCNs and 746 IPMNs resected at a single institution. MCNs were reviewed for confirmation of ovarian-type stroma and reclassified according to current WHO guidelines. Results: MCNs presented almost exclusively in middle-aged women (median 47.5 years, 96.5% female) as solitary (100%), macrocystic (94.2%) lesions in the distal pancreas (92.1%). IPMNs were distributed equally by sex in an older population (median 69.0 years, 49.6% female) and favored the proximal pancreas (67.6%). Compared with IPMNs, MCNs were larger (4.2 cm vs 2.5 cm) and more often low-grade (71.1% vs 13.8%). Associated invasive carcinoma was less common in MCNs than in IPMNs (9.9% vs 32.4%). Surgical resection was curative for 100% of noninvasive MCNs. Patients with an MCN-associated invasive carcinoma had a much better prognosis than did patients with an IPMN-associated invasive carcinoma with 10-year disease-specific survival of 79.6% versus 27.2%, respectively. Conclusion: MCNs have a stereotypical clinical profile that is readily distinguishable from IPMNs based on demographic features, imaging, and pathology. Most MCNs are noninvasive and curable with surgical resection. Prognosis remains excellent even for invasive disease with 10-year survival approaching 80% following resection. © 2017 IAP and EPC. Published by Elsevier B.V. All rights reserved.
Keywords: Pancreatic mucinous cystic neoplasm (MCN) Ovarian stroma Pancreatic cystic neoplasms Intraductal papillary mucinous neoplasm (IPMN) Pancreatic adenocarcinoma (PDAC)
Introduction Mucinous cystic neoplasms (MCNs) and intraductal papillary mucinous neoplasms (IPMNs) are cystic precursors to invasive ductal adenocarcinoma (PDAC) characterized by a neoplastic
* Corresponding author. Department of Surgery, The Johns Hopkins University School of Medicine, Halsted 608, 600 N Wolfe Street, Baltimore, MD, 21287, USA. E-mail address: [email protected] (M.J. Weiss).
mucin-producing epithelial lining. MCNs were originally described more than a century ago, but their significance wasn't fully appreciated until 1978 when Compagno and Oertel emphasized the malignant potential of mucin-producing cystic neoplasms over their serous counterparts [1]. The first description of IPMNs followed just a few years later in 1982 when Ohashi et al. reported four cases of a distinctive “mucus-secreting pancreatic cancer” originally thought to be a subtype of MCN disease rather than a separate neoplasm [2,3].
http://dx.doi.org/10.1016/j.pan.2017.04.003 1424-3903/© 2017 IAP and EPC. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
2
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
MCNs are distinguished from IPMNs by the presence of a distinctive ovarian-type stroma comprised of densely packed spindle-cells that may express progesterone and estrogen receptors [4,5]. MCNs generally do not communicate with the ductal system and occur almost exclusively in the body and tail regions of the pancreas. They tend to be quite large with a median size of 4e5 cm, though some have been documented up to 35 cm. MCNs have a strong female preponderance (20:1) and are usually diagnosed in the fourth and fifth decades of life. IPMNs lack ovarian-type stroma and by definition involve the ductal system. They affect both sexes equally between the sixth and seventh decades of life and affect the head of the pancreas more often than the tail [6]. The World Health Organization (WHO) finally recognized IPMN disease as a distinct histopathological entity in 1996 [7], but ambiguous diagnostic criteria and terminology resulted in ongoing misclassification of branch-duct type IPMNs as MCNs. It was not until 2006 that the diagnostic criteria were revised to include ovarian-type stroma as a distinguishing feature of MCNs [8]. However, by this point, the growing body of MCN literature consisted mostly of flawed studies contaminated with IPMNs. The goal of the present study is to clarify the clinical features, malignant potential, and long-term outcomes of MCNs defined by the presence of ovarian stroma in the largest single-institution, pathologyconfirmed series to date.
Methods Patients After obtaining approval from the Johns Hopkins Institutional Review Board, cases were identified from a prospectively managed pathology database. All potential MCNs were reviewed under the direct supervision of a single senior pathologist specializing in pancreatic pathology (RHH). Study inclusion was contingent upon the identification of ovarian stroma and confirmed cases were regraded according to the 2010 WHO guidelines [9]. A comparison cohort of resected IPMNs was also identified consisting of cases resected after the 1996 WHO classification through 2015. Reresections, cases with an unspecified grade of dysplasia, and IPMNs diagnosed secondary to an unrelated primary malignancy were all excluded. Clinical data for all study subjects were collected from the electronic medical record and original paper charts.
Analysis Statistical analyses were performed using the Stata 13 statistical software package (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP) and GraphPad Prism version 7.00 for Mac (GraphPad Software, La Jolla California USA, www.graphpad.com). All datasets were non-normally distributed based on the Shapiro-Wilk W test and nonparametric tests of significance were used for analyses. All analyses were performed as two-tailed tests with a significance threshold of a ¼ 0.05. Categorical data were expressed as relative frequencies (%) with total observations (n) and evaluated using Fisher's exact test or the chi-square test. Continuous data were summarized as medians with interquartile ranges (IQR) and compared using the Wilcoxon or Kruskal-Wallis rank tests depending on the number of comparison groups. For matched datasets, analogous evaluations were performed using either the Wilcoxon signed-rank test, McNemar test, or Stuart-Maxwell marginal homogeneity test. The last is a generalization of the McNemar test that allows for comparison of multilevel categorical variables in matched pairs.
Survival analysis Overall (OS) and disease-specific survival (DSS) were assessed by the Kaplan-Meier method with log-rank comparisons between cohorts and subgroups of interest. Survival was defined as the total time in years between resection and death or date of censoring. DSS includes only those deaths attributed to pancreatic cancer or its complications. Follow-up time and survival were calculated using the most recent confirmed medical encounter based on our records. Patients without confirmed deaths were not assumed to be alive at the time of publication and were censored at their most recent follow-up. Deaths were determined by chart review and through the Social Security Death Index (SSDI) with causes obtained from the medical record and death certificates. Results Demographics and clinicopathological features From 1984 through 2015, 142 pathology-confirmed MCNs were resected at the Johns Hopkins Hospital, accounting for <2% of all pancreatic resections during that timeframe. Of 142 MCNs, 101 (71.1%) had low-grade dysplasia (LGD), 15 (10.6%) intermediategrade dysplasia (IGD), 12 (8.5%) high-grade dysplasia (HGD), and 14 (9.9%) had an associated invasive cancer (ICA). The IPMN cohort consisted of 746 cases resected between 1996 and 2015. Pathology results included 103 (13.8%) cases with LGD, 247 (33.1%) with IGD, 154 (20.6%) with HGD, and 242 (32.4%) with an associated ICA. Detailed patient demographics and clinicopathological data are presented in Table 1. Males with an MCN were extremely uncommon in our series (n ¼ 5, 3.5%) and tended to be much older than their female counterparts [59.0 (56.0e61.0) years vs 47.0 (37.0e57.0) years, p ¼ 0.03]. Among female patients, 66 (48.2%) were premenopausal and 71 (51.8%) were peri- (n ¼ 6) or postmenopausal (n ¼ 65). Ten (14.9%) premenopausal women presented during or shortly after pregnancy and 16 (22.5%) peri/postmenopausal patients reported a history of estrogen-based HRT. No baseline demographics evaluated in this study were associated with the pathological finding of invasive disease. However, MCNs with an associated ICA were significantly larger on pathology compared to those without an associated ICA (p ¼ 0.0002). As demonstrated by Fig. 1, increasing tumor size was also associated with increasing grade of dysplasia (p ¼ 0.0001). Clinical features in IPMNs associated with the presence of ICA included advanced age (p ¼ 0.003), decreased body mass index (BMI) (p ¼ 0.001), history of heavy alcohol use (p ¼ 0.04), preoperative diabetes mellitus (p ¼ 0.003), and American Society of Anesthesia (ASA) class (p < 0.001). For the purpose of this study, “heavy alcohol use” was defined as a documented history of alcoholic pancreatitis, alcohol dependence, or reported consumption at or approaching daily use. Surgical resections for IPMNs with an associated ICA consisted of more TPs and fewer DPs compared to IPMNs without an associated ICA (p < 0.001). Invasive disease was also associated with greater lymph node yield (p < 0.0001) and larger tumor size (p < 0.001) on pathology. MCN presenting symptoms Patients were considered symptomatic if complaints were consistent with a pancreatic source defined as left upper quadrant (LUQ) or epigastric pain in the absence of another obvious and identifiable causes (e.g., cholelithiasis, diverticulitis, nephrolithiasis). Abdominal pain was the most common presenting complaint (n ¼ 93, 65.5%), but only 72 (50.7%) had pancreas-type
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
3
Table 1 Demographics and clinicopathological features of the study population. Baseline Demographics
MCN All
IPMN Invasive
Noninvasive
Pc
MCN vs IPMN
All
Invasive
Noninvasive
Pd
Pe
Total Subjects, (n)
142
14
128
746
242
504
Age, median (IQR)
47.5 (37.0 e57.0)
42.0 (36.8 e57.5)
48.0 (37.0 e57.0)
0.68b
69.0 (61.0 e75.0)
70.0 (62.0 e77.0)
68.0 (60.0 e74.0)
0.003b
<0.0001b
Female, % (n):
96.5% (137)
92.9% (13)
96.9% (124)
0.41a
49.6% (370)
44.6% (108)
52.0% (262)
0.06a
<0.001a
Race, % (n): Caucasian African American Asian Other
71.8% (102) 18.3% (26) 4.9% (7) 4.9% (7)
92.9% (13) 7.1% (1) 0.0% (0) 0.0% (0)
69.5% (89) 19.5% (25) 5.5% (7) 5.5% (7)
0.19a 0.31a 1.0a 1.0a
89.0% (664) 5.4% (40) 2.1% (16) 3.5% (26)
90.5% (219) 5.8% (14) 1.2% (3) 2.5% (6)
88.3% (445) 5.2% (26) 2.6% (13) 4.0% (20)
0.52a 0.73a 0.29a 0.63a
<0.001a <0.001a 0.07a 0.23a
BMI, median (IQR)
27.6 (23.6 e30.9)
29.1 (24.3 e30.6)
27.5 (23.1 e32.3)
0.71b
25.7 (22.8 e29.0)
24.6 (22.0 e27.2)
26.4 (23.1 e29.5)
0.001b
0.001b
Smoking History, % (n):
47.9% (68)
64.3% (9)
46.1% (59)
0.27a
56.9% (413)
56.2% (136)
55.0% (277)
0.69a
0.11a
Heavy Alcohol Use, % (n):
12.7% (18)
14.3% (2)
12.5% (16)
1.0
a
19.5% (126)
22.3% (54)
14.3% (72)
0.04
a
a
22.3% (166)
28.9% (70)
19.0% (96)
0.003
0.09a a
<0.001a
Diabetes Mellitus, % (n):
7.7% (11)
0.0% (0)
8.6% (11)
0.60
ASA Class, % (n): I II III IV
3.8% (4) 61.9% (65) 34.3% (36) 0% (0)
7.1% (1) 35.7% (5) 35.7% (5) 0.0% (0)
2.3% (3) 46.9% (60) 24.2% (31) 0.0% (0)
0.21a
0.2% (1) 32.6% (210) 65.5% (422) 1.7% (11)
0.0% (0) 20.2% (49) 62.4% (151) 3.3% (8)
0.2% (1) 31.9% (161) 53.8% (271) 0.6% (3)
<0.001a
<0.001a
Extra-Pancreatic Malignancy, Any Breast (% females) Prostate (% males)
% (n): 12.0% (17) 6.6% (9) 20% (1)
21.4% (3) 15.4% (2) 0% (0)
10.2% (13) 5.6% (7) 25% (1)
0.20a
21.0% (157) 7.6% (28) 10.4% (39)
19.0% (46) 1.9% (2) 10.4% (14)
22.0% (111) 9.9% (26) 10.3% (25)
0.50a
0.01a
78.6% (11) 14.3% (2) 7.1% (1) 0.0% (0) 0.0% (0)
91.4% (117) 4.7% (6) 0.8% (1) 0.8% (1) 2.3% (3)
0.15a
24.3% (181) 65.1% (486) 8.7% (65) 1.3% (10) 0.5% (4)
17.0% (41) 64.3% (155) 18.7% (45) 0.0% (0) 0.0% (0)
27.8% (140) 65.4% (329) 4.0% (20) 2.0% (10) 0.8% (4)
<0.001a
<0.001a
Surgical Resections & Pathology Resection Type, % (n): DP PD TP CP Enucleation
90.1% (128) 5.6% (8) 1.4% (2) 0.7% (1) 2.1% (3)
Tumor Size (cm), median (IQR) 4.2 (2.5e7.0)
10.0 (4.9e14.0) 4.0 (2.5e6.4)
0.0002b 2.5 (1.5e3.5)
3.0 (1.9e5.2)
2.3 (1.5e3.0)
<0.0001b <0.0001b
Nodal Status: Total LN, Median (IQR) Positive LN, % (n)
13 (6e21.3) 0.7% (1)
16 (10.5e25) 7.1% (1)
12.5 (6e21) 0.0% (0)
0.23b
17 (13e23) 16.5% (123)
21 (14e28) 50.8% (123)
15 (12e21) 0.0% (0)
<0.0001b <0.0001b 0.001a
Invasion, % (n): Lymphovascular Perineural
0.7% (1) 1.4% (2)
7.1% (1) 14.3% (2)
0.0% (0) 0.0% (0)
10.9% (81) 20.0% (149)
33.5% (81) 61.6% (149)
0.0% (0) 0.0% (0)
0.004a <0.001a
Significant p-values (<0.05) are in bold. Note. BMI, body mass index; ASA, America Society of Anesthesiologists; IQR, interquartile range; MPD, main pancreatic duct; DP, distal pancreatectomy; PD, pancreaticoduodenectomy; TP, total pancreatectomy; CP, central pancreatectomy; LN, lymph node; MCN, mucinous cystic neoplasm; P, p-value; IPMN, intraductal papillary mucinous neoplasm. a Fisher's Exact; b Mann-Whitney U. c Comparison of invasive vs noninvasive MCN. d Comparison of invasive vs noninvasive IPMN. e Comparison of total MCN vs total IPMN.
pain. Additional complaints included pain radiating to the back (n ¼ 30, 21.1%), nausea (n ¼ 33, 23.2%), post-prandial symptoms such as worsening pain and early satiety (n ¼ 26, 18.3%), vague gastrointestinal complaints including crampiness, gas, or bloating (n ¼ 24, 16.9%), weight loss (n ¼ 20, 14.1%), and diarrhea (n ¼ 14, 9.9%). Additionally, 28 (19.7%) were diagnosed with acute pancreatitis and 21 (14.8%) had a palpable abdominal mass on physical exam. The presence of an associated ICA was associated with a palpable mass on exam (p < 0.001) and postprandial symptoms (p ¼ 0.005). Tumor size was associated with a palpable mass and post-prandial symptoms as well (p < 0.0001 and p ¼ 0.002,
respectively), along with nausea (p ¼ 0.014 and diarrhea (p ¼ 0.031). Abdominal pain demonstrated no significant associations.
Diagnostic imaging analysis Imaging reports from preoperative computed tomography (CT) were reviewed for 139 MCN patients and 457 IPMN patients (Table 2). For patients with an MCN, CT findings associated with ICA included increasing tumor size (p ¼ 0.0003), mural nodules (p ¼ 0.004), internal septations (p ¼ 0.03), and solid or solid-cystic features in place of the typical macrocystic features (p ¼ 0.004).
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
4
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
Fig. 1. Comparison of MCN and IPMN tumor size by grade of dysplasia. Tumor size was obtained from computed tomography (CT) scans (left) and pathology (right). Points correspond to median (IQR) tumor sizes, which are recorded beneath the corresponding grade of dysplasia. P-values were obtained from Mann-Whitney U comparison of MCN and IPMN tumor size at each grade of dysplasia. LGD ¼ low-grade dysplasia, IGD ¼ intermediate-grade dysplasia, HGD ¼ high-grade dysplasia, and ICA ¼ invasive carcinoma.
Table 2 CT imaging features. CT Imaging Features
MCN
IPMN
MCN vs IPMN
All
Invasive
Noninvasive
P
All
Invasive
Noninvasive
P
Pe
Size (cm): Median (IQR)
4.0 (3.0e6.6)
10.0 (5.0e16.1)
3.8 (3.0e6.1)
0.0003b
3.0 (2.0e4.0)
3.3 (1.9e4.7)
2.9 (2.0e3.8)
0.047b
<0.0001b
Location, % (n): Proximal Distal Proximal & Distal
7.9% (11) 92.1% (128) 0% (0)
21.4% (3) 78.6% (11) 0% (0)
6.4% (8) 93.6% (117) 0% (0)
0.08a 0.08a e
67.6% (309) 24.9% (114) 7% (32)
68.2% (88) 24.8% (32) 7% (9)
67.4% (221) 25% (82) 7% (23)
0.91a 1.0a 1.0a
<0.001a <0.001a <0.001a
100% (139) 94.2% (131) 9.4% (13)
100% (14) 71.4% (10) 35.7% (5)
100% (125) 96.8% (121) 6.4% (8)
e 0.004a 0.004a
64.8% (296) 70.2% (321) 22.1% (101)
61.2% (79) 51.2% (66) 41.1% (53)
57% (187) 77.7% (255) 14.6% (48)
0.83a <0.001a <0.001a
<0.001a <0.001a 0.001a
38.1% (53) 12.9% (18) 2.9% (4) 0% (0)
64.3% (9) 0% (0) 14.3% (2) 0% (0)
35.2% (44) 14.4% (18) 1.6% (2) 0% (0)
0.04a 0.22a 0.05a e
37.2% (170) 5.5% (25) 46.4% (212) 68.7% (314)
32.6% (42) 6.2% (8) 69.8% (90) 62% (80)
39% (128) 5.2% (17) 37.5% (123) 71.6% (235)
0.24a 0.65a <0.001a 0.06a
0.84a 0.005a <0.001a <0.001a
Features, % (n): Solitary Lesion Macrocystic Mural Nodule/Solid Component Size (mm) Septations Calcifications Dilated MPD MPD Communication
c
d
Significant p-values (<0.05) are in bold. Note. CT, computed tomography; IQR, interquartile range; MPD, main pancreatic duct; MCN, mucinous cystic neoplasm; P, p-value; IPMN, intraductal papillary mucinous neoplasm. a Fisher's Exact; b Mann-Whitney U. c Comparison of invasive vs noninvasive MCN. d Comparison of invasive vs noninvasive IPMN. e Comparison of total MCN vs total IPMN.
CT evidence of splenic vein thrombosis was also strongly correlated with MCN-associated ICA. After excluding proximal MCNs, there were 98 cases (72 LGD, 10 IGD, 8 HGD, 8 ICA) with CT reports commenting on splenic vein (SV) patency. Thrombosis was observed in all 8 (100%) MCNs with an associated ICA compared with only 10 (11.1%) out of 90 MCNs without ICA (p<0.001). For those with an IPMN, CT findings associated with ICA included increasing tumor size (p ¼ 0.047), mural nodules (p < 0.001), absence of macrocystic features (p < 0.001), and main pancreatic duct (MPD) dilation (p < 0.001).
Disease-specific survival Detailed DSS data are presented in Table 3 with corresponding Kaplan-Meier curves in Fig. 2. There were only 2 disease-specific deaths in the entire MCN cohort; both occurred in the setting of an associated ICA, 1 of which had positive lymph nodes. Absence of an associated ICA demonstrated DSS of 100% at all time points, while presence of an associated ICA (all cases) had DSS of 90.9%, 79.6%, and 79.6% at 1, 5, and 10 years, respectively. Considering only N0 invasive disease, DSS was 100%, 87.5%, and 87.5%, respectively. For the IPMN cohort, disease-specific deaths were confirmed in 17 (3.4%) patients without an associated ICA, 124 (51.2%) patients
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
5
Table 3 Summary results of Kaplan-Meier survival analyses. Full Cohort
Invasive (N0 Only)
Noninvasive
MCN
IPMN
MCN
Invasive (All) IPMN
MCN
IPMN
MCN
IPMN
Overall Survival At risk, (n) Median (IQR), y 95% CI Deaths, % (n) 1-year, % 5-year, % 10-year, % Log-Rank (P)
140 e e 7.1% (10) 98.9% 88.1% 80.9% <0.0001
746 7.8 (2.9e13.6) 6.7e8.9 32.8% (245) 91.0% 62.3% 37.5%
14 e e 14.3% (2) 90.9% 79.6% 79.6% 0.0003
242 2.4 (1.2e6.4) 2.0e3.2 64.0% (155) 79.2% 31.2% 12.5%
12 e e 8.3% (1) 100% 87.5% 87.5% 0.002
118 5.4 (2.0e8.8) 3.8e7.7 51.7% (61) 86.9% 51.1% 22.0%
126 e e 6.3% (8) 100% 89.6% 80.4% 0.0008
504 10.9 (6.8e15.4) 9.8e13.8 17.8% (90) 97.4% 81.6% 54.8%
Disease-Specific At risk, (n) Median (IQR), y 95% CI Deaths, % (n) 1-year, % 5-year, % 10-year, % Log-Rank (P)
140 e e 1.4% (2) 98.9% 96.8% 96.8% <0.0001
746 e e 18.9% (141) 92.7% 72.9% 63.1%
14 e e 14.3% (2) 90.9% 79.6% 79.6% 0.01
242 2.7 (1.3-.) 2.0e3.8 51.2% (124) 80.3% 36.0% 27.2%
12 e e 8.3% (1) 100% 87.5% 87.5% 0.058
118 7.8 (2.0-.) 2.2e4.4 33.9% (40) 88.7% 57.7% 44.3%
126 e e 0% (0) 100% 100% 100% 0.052
504 e e 3.4% (17) 99.5% 95.9% 86.2%
Significant p-values (<0.05) are in bold. Note. IQR, interquartile range; CI, confidence interval; P, p-value; MCN, mucinous cystic neoplasm; IPMN, intraductal papillary mucinous neoplasm.
with an associated ICA, and 40 (33.9%) with N0 ICA. DSS at 1, 5, and 10 years was 99.5%, 95.9%, and 86.2%, respectively for noninvasive IPMNs, 80.3%, 36.0%, and 36.0%, respectively for those with an associated ICA (all cases), and 88.7%, 57.7%, and 44.3% respectively for N0 invasive disease. The log-rank test demonstrated significantly better DSS for patients with MCNs compared to those with IPMNs overall (p < 0.0001) and in the setting of an associated ICA (p ¼ 0.02), while N0 disease (p ¼ 0.58) and noninvasive disease (p ¼ 0.052) approached significance. Overall survival Detailed OS data, including subanalyses by invasiveness, are presented in Table 3 with corresponding Kaplan-Meier curves in Fig. 2. There were 10 (7.0%) deaths overall in the MCN cohort. Eight occurred in patients with noninvasive disease from causes unrelated to their pancreatic neoplasms, while the other two patients died from MCN-associated ICA. OS at 1, 5, and 10 years was 98.9%, 88.1%, and 80.9%, respectively. There were 245 (32.8%) total deaths recorded in the IPMN cohort, 90 of which occurred in patients with noninvasive disease and 155 in patients with an IPMN with associated ICA. Median OS for the entire IPMN cohort was 7.8 (2.9e13.6) years with 1-, 5-, and 10-year OS of 91.0%, 62.3%, and 37.5%, respectively. Log-rank comparison confirmed significantly greater OS for patients with an MCN compared to those with an IPMN (p < 0.0001). OS in the subsets of noninvasive and invasive disease also demonstrated significantly higher survival for MCNs compared to IPMNs over the 1-, 5-, and 10-year intervals (Table 3). Discussion This study was conducted at a high-volume center for pancreatic diseases and comprises the largest single-institution series of pathology-confirmed MCNs to date. Based on our experience, MCNs have a highly stereotypical clinical profile presenting almost exclusively in middle-aged women as solitary, macrocystic lesions in the body and tail of the pancreas. By comparison, IPMNs usually present as proximal tumors in older patients with a nearly equal sex distribution. Survival analysis demonstrated a much better prognosis for
patients with an MCN compared to those with an IPMN due to the infrequent progression of MCNs to ICA and better survival once ICA develops. The overwhelming majority of MCNs presented with noninvasive disease, all of which were cured by surgical resection. However, a more interesting observation is that patients with MCNassociated ICA also had a favorable prognosis, particularly compared to the 51.2% disease-specific mortality rate of patients with IPMN-associated ICA. Only 2 (14.3%) of the 14 patients with an MCN-associated ICA died, and all 12 survivors remained alive without evidence of recurrence at the time of this study. Included among these were several examples of long-term survival exceeding 10 years (n ¼ 3) and 20 years (n ¼ 2) with an overall median follow-up time of 42.8 (6.4e166.7) months. Survival for patients with a surgically resected MCN or IPMN was closely correlated with pathology results, particularly lymph node involvement, lymphovascular invasion, and perineural invasion. These are high-risk pathologic features that signal the presence of more advanced, likely micrometastatic disease associated with a high rate of recurrence and poor prognosis [10,11]. Though very common among IPMN-associated ICA, high-risk pathologic features were rarely identified in MCN-associated ICA, which likely contributes to their favorable prognosis. Comparable results were reported by Crippa et al. in 19 MCNs with associated ICA [12]. Similarly, a recent study from this institution suggested that the presence of minimally-invasive disease limited to the ovarian stroma is not a poor prognosticator in MCNs [13]. Recent studies have suggested the possibility of nonoperative management for small (<3e4 cm) MCNs without high-risk features, particularly in borderline surgical candidates [4,12]. In our series, the smallest MCN-associated ICA measured 4.0 cm on CT. There were 41 (28.9%) MCNs measuring <4 cm without mural nodules, including 40 with LGD and a single MCN with IGD. These results suggest that imaging criteria can effectively identify lowgrade lesions and close imaging surveillance may be a reasonable management option for small (<4 cm) MCN without concerning imaging features. Furthermore, since only 2 (14.3%) of the 14 patients with MCN-associated ICA died from their disease, even those progressing during surveillance may still have a very good prognosis. However, since MCNs generally present in otherwise young and healthy patients, “lifelong surveillance” could entail several
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
6
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
Fig. 2. Kaplan-Meier survival curves depicting survival among MCN (dashed) and IPMN (solid). I.) Overall and disease-specific survival in the full MCN and IPMN cohorts. Log-rank comparison demonstrated significantly better overall (A) and disease-specific (B) survival for MCN compared to IPMN (p < 0.0001 and p < 0.0001, respectively). II.) Disease-specific survival in MCN and IPMN with (C) and without (D) an associated ICA. Log-rank comparison demonstrated significantly better disease-specific survival for MCN-associated ICA versus IPMN-associated ICA overall (p ¼ 0.02) and approached significance when comparing only N0 disease (p ¼ 0.058). Log-rank comparison of MCN and IPMN without an associated ICA also approached significance (p ¼ 0.052).
decades of imaging follow-up and most patients present as excellent surgical candidates (65.7% ASA I/II, Table 1). Therefore, when considering nonoperative management, surgeons must weigh the associated factors of cost-effectiveness, radiation exposure, compliance with very long-term follow-up, and potential for disease progression and death against the fact that our series demonstrated resection to be curative for all cases of non-invasive disease with no operative mortality. The overwhelming likelihood is that most patients do not require surgical management, but there is currently no long-term data available to suggest which patients are at risk for progression to ICA. Therefore, larger series and better long-term data are required to identify those patients who would benefit most from surgery. The vast majority of MCNs arise in women and some studies have linked them to pregnancy and hormone-replacement therapy
(HRT) [14e19]. As a result, some have speculated that MCN may be hormonally influenced. In our series, we did not observe any statistically significant associations with pregnancy or HRT. However, we made the interesting observation that premenopausal women accounted 9/13 (69.2%) female MCN-associated ICAs. If validated in other studies, this trend would support the notion of hormonal influence on tumor progression and suggest that premenopausal patients are at greatest risk for cancer. Furthermore, a reduction in risk following menopause could have significant implications for current management recommendations. As it currently stands, this is an isolated, nonsignificant finding and requires additional study in a much larger series. As with any retrospective study, our data were limited to data available in the medical record. The fact that patients often originated as referrals from other institutions meant that records were
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
often limited for data unrelated to their primary pancreatic disease. These patients usually transitioned care closer to home following surgery, which also decreased the availability of long-term data and follow-up. This is particularly true for MCN patients, who often did not require long-term surgical follow-up. One final limitation of the study was the rarity of MCNs, which greatly limits the size of a single institution series.
[6] [7]
[8]
Conclusion [9]
This is the largest single-institution series of MCNs reported to date with 142 cases confirmed by the presence of ovarian stroma on pathology review. Based on our results, MCNs are rare tumors with a consistent presentation and infrequent progression to cancer. For MCNs without an associated cancer, complete surgical resection was curative in 100% of cases. For those with MCN-associated ICA, prognosis was still excellent with a 14.3% mortality rate and 10-year survival approaching 80%.
[10]
[11]
[12]
Acknowledgements Supported by grants from the National Cancer Institute of the National Institutes of Health (T32CA126607 and CA62824), the Michael Rolfe Foundation, and the Sol Goldman Pancreatic Cancer Research Center. The authors have no conflicts of interest to disclose.
[13]
[14]
[15]
References [16] [1] Compagno J, Oertel JE. Mucinous cystic neoplasms of the pancreas with overt and latent malignancy (cystadenocarcinoma and cystadenoma). A clinicopathologic study of 41 cases. Am J Clin Pathol 1978;69:573e80. [2] Ohashi K, Murakami Y, Takeoshi T, Ohta HOI. Four cases of mucin producing cancer of the pancreas on specific findings of the papilla of Vater. Prog Dig Endosc 1982;20:348e51. [3] Itai Y, Ohhashi K, Nagai H, Murakami Y, Kokubo T, Makita K, et al. “Ductectatic” mucinous cystadenoma and cystadenocarcinoma of the pancreas. Radiology 1986;161:697e700. http://dx.doi.org/10.1148/radiology.161.3.3786719. [4] Reddy RP, Smyrk TC, Zapiach, M, Levy MJ, Pearson RK, Clain JE, et al. Pancreatic mucinous cystic neoplasm defined by ovarian stroma: demographics, clinical features, and prevalence of cancer. Clin Gastroenterol Hepatol 2004;2: 1026e31. http://dx.doi.org/10.1016/S1542-3565(04)00450-1. [5] de Wilde RF, Hruban RH, Maitra A, Offerhaus GJA. Reporting precursors to
[17]
[18]
[19]
7
invasive pancreatic cancer: pancreatic intraepithelial neoplasia, intraductal neoplasms and mucinous cystic neoplasm. Diagn Histopathol 2012;18:17e30. http://dx.doi.org/10.1016/j.mpdhp.2011.10.012. Lennon AM, Wolfgang C. Cystic neoplasms of the pancreas. J Gastrointest Surg 2013;17:645e53. http://dx.doi.org/10.1007/s11605-012-2072-6. Kloppel G, Solicia E, Longnecker D, Capella C, Sobin L. Histologic typing of tumors of the exocrine pancreas: World Health organization international histologic classification of tumors. second ed. Berlin: Springer-Verlag; 1996. Tanaka M, Chari S, Adsay V, Fernandez-del Castillo C, Falconi M, Shimizu M, et al. International consensus guidelines for management of intraductal papillary mucinous neoplasms and mucinous cystic neoplasms of the pancreas. Pancreatology 2006;6:17e32. http://dx.doi.org/10.1159/000090023. Zamboni G, Fukushima N, Hruban RH. Mucinous cystic neoplasms of the pancreas. In: Bosman FE, Carneiro F, Hruban H, Theise N, editors. WHO Classif. Tumors Dig. Syst. fourth ed. Lyon: IARC; 2010. p. 300e3. Winter JM, Cameron JL, Campbell KA, Arnold MA, Chang DC, Coleman J, et al. 1423 pancreaticoduodenectomies for pancreatic cancer: a single-institution experience. J Gastrointest Surg 2006;10:1199e210. http://dx.doi.org/10.1016/ j.gassur.2006.08.018. Nathan H, Wolfgang CL, Edil BH, Choti MA, Herman JM, Schulick RD, et al. Perioperative mortality and long-term survival after total pancreatectomy for pancreatic adenocarcinoma: a population-based perspective. J Surg Oncol 2009;99:87e92. http://dx.doi.org/10.1002/jso.21189. Crippa S, Salvia R, Warshaw AL, Domínguez I, Bassi C, Falconi M, et al. Mucinous cystic neoplasm of the pancreas is not an aggressive entity: lessons from 163 resected patients. Ann Surg 2008;247:571e9. http://dx.doi.org/ 10.1097/SLA.0b013e31811f4449. Lewis GH, Wang H, Bellizzi AM, Klein AP, Askin FB, Schwartz LE, et al. Prognosis of minimally invasive carcinoma arising in mucinous cystic neoplasms of the pancreas. Am J Surg Pathol 2013;37:601e5. http://dx.doi.org/10.1097/ PAS.0b013e318273f3b0. Kosumi K, Takamori H, Hashimoto D, Tanaka H, Abe S, Nakahara O, et al. Mucinous cystic neoplasm of the pancreas activated during pregnancy. Surg Case Rep 2015;1:13. http://dx.doi.org/10.1186/s40792-014-0012-2. Tica AA, Tica OS, Saftoiu A, Camen D, Tica VI. Large pancreatic mucinous cystic neoplasm during pregnancy: what should be done? Gynecol Obstet Invest 2013;75:132e8. http://dx.doi.org/10.1159/000346176. Tsuda H, Kotani T, Sumigama S, Mano Y, Shimoyama Y, Kikkawa F. Mucinous cystic neoplasm of the pancreas with severe dysplasia during pregnancy: case report and review of the literature. Taiwan J Obstet Gynecol 2012;51:635e8. http://dx.doi.org/10.1016/j.tjog.2012.09.022. Kosumi K, Takamori H, Hashimoto D, Tanaka H, Abe S, Nakahara O, et al. Mucinous cystic neoplasm of the pancreas activated during pregnancy. Surg Case Rep 2015;1:13. http://dx.doi.org/10.1186/s40792-014-0012-2. Tanaka S, Kawamura T, Nakamura N, Teramoto K, Arii S. Mucinous cystadenocarcinoma of the pancreas developing during hormone replacement therapy. Dig Dis Sci 2007;52:1326e8. http://dx.doi.org/10.1007/s10620-0069482-9. Foster D, Shaikh MF, Gleeson E, Babcock BD, Lin J, Ownbey RT, et al. Pancreatic mucinous cystic neoplasm in a transgender patient. World J Surg Oncol 2015;13:205. http://dx.doi.org/10.1186/s12957-015-0620-8.
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
Contents lists available at ScienceDirect
Pancreatology journal homepage: www.elsevier.com/locate/pan
Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series James F. Griffin, MD a, Andrew J. Page, MD a, Georges J. Samaha, MD a, Adrienne Christopher, BA a, Feriyl Bhaijee, MD b, Maryam K. Pezhouh, MD b, Niek A. Peters, BSc a, Ralph H. Hruban, MD b, Jin He, MD a, Martin A. Makary, MD a, Anne Marie Lennon, MD, PhD c, John L. Cameron, MD a, Christopher L. Wolfgang, MD, PhD a, Matthew J. Weiss, MD a, * a b c
Departments of Surgery, The Johns Hopkins University School of Medicine, Baltimore, MD, USA Departments of Pathology, The Johns Hopkins University School of Medicine, Baltimore, MD, USA Departments of Gastroenterology, The Johns Hopkins University School of Medicine, Baltimore, MD, USA
a r t i c l e i n f o
a b s t r a c t
Article history: Received 19 November 2016 Received in revised form 16 March 2017 Accepted 7 April 2017 Available online xxx
Background/Objectives: Mucinous cystic neoplasms (MCNs) are rare pancreas tumors distinguished from intraductal papillary mucinous neoplasms (IPMNs) by the presence of ovarian-type stroma. Historical outcomes for MCNs vary due to previously ambiguous diagnostic criteria resulting in confusion with IPMNs. This study seeks to characterize and clarify the clinical features and long-term outcomes of MCNs versus IPMNs in the largest single-institution series of pathology-confirmed MCNs to date. Methods: We compared 142 MCNs and 746 IPMNs resected at a single institution. MCNs were reviewed for confirmation of ovarian-type stroma and reclassified according to current WHO guidelines. Results: MCNs presented almost exclusively in middle-aged women (median 47.5 years, 96.5% female) as solitary (100%), macrocystic (94.2%) lesions in the distal pancreas (92.1%). IPMNs were distributed equally by sex in an older population (median 69.0 years, 49.6% female) and favored the proximal pancreas (67.6%). Compared with IPMNs, MCNs were larger (4.2 cm vs 2.5 cm) and more often low-grade (71.1% vs 13.8%). Associated invasive carcinoma was less common in MCNs than in IPMNs (9.9% vs 32.4%). Surgical resection was curative for 100% of noninvasive MCNs. Patients with an MCN-associated invasive carcinoma had a much better prognosis than did patients with an IPMN-associated invasive carcinoma with 10-year disease-specific survival of 79.6% versus 27.2%, respectively. Conclusion: MCNs have a stereotypical clinical profile that is readily distinguishable from IPMNs based on demographic features, imaging, and pathology. Most MCNs are noninvasive and curable with surgical resection. Prognosis remains excellent even for invasive disease with 10-year survival approaching 80% following resection. © 2017 IAP and EPC. Published by Elsevier B.V. All rights reserved.
Keywords: Pancreatic mucinous cystic neoplasm (MCN) Ovarian stroma Pancreatic cystic neoplasms Intraductal papillary mucinous neoplasm (IPMN) Pancreatic adenocarcinoma (PDAC)
Introduction Mucinous cystic neoplasms (MCNs) and intraductal papillary mucinous neoplasms (IPMNs) are cystic precursors to invasive ductal adenocarcinoma (PDAC) characterized by a neoplastic
* Corresponding author. Department of Surgery, The Johns Hopkins University School of Medicine, Halsted 608, 600 N Wolfe Street, Baltimore, MD, 21287, USA. E-mail address: [email protected] (M.J. Weiss).
mucin-producing epithelial lining. MCNs were originally described more than a century ago, but their significance wasn't fully appreciated until 1978 when Compagno and Oertel emphasized the malignant potential of mucin-producing cystic neoplasms over their serous counterparts [1]. The first description of IPMNs followed just a few years later in 1982 when Ohashi et al. reported four cases of a distinctive “mucus-secreting pancreatic cancer” originally thought to be a subtype of MCN disease rather than a separate neoplasm [2,3].
http://dx.doi.org/10.1016/j.pan.2017.04.003 1424-3903/© 2017 IAP and EPC. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
2
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
MCNs are distinguished from IPMNs by the presence of a distinctive ovarian-type stroma comprised of densely packed spindle-cells that may express progesterone and estrogen receptors [4,5]. MCNs generally do not communicate with the ductal system and occur almost exclusively in the body and tail regions of the pancreas. They tend to be quite large with a median size of 4e5 cm, though some have been documented up to 35 cm. MCNs have a strong female preponderance (20:1) and are usually diagnosed in the fourth and fifth decades of life. IPMNs lack ovarian-type stroma and by definition involve the ductal system. They affect both sexes equally between the sixth and seventh decades of life and affect the head of the pancreas more often than the tail [6]. The World Health Organization (WHO) finally recognized IPMN disease as a distinct histopathological entity in 1996 [7], but ambiguous diagnostic criteria and terminology resulted in ongoing misclassification of branch-duct type IPMNs as MCNs. It was not until 2006 that the diagnostic criteria were revised to include ovarian-type stroma as a distinguishing feature of MCNs [8]. However, by this point, the growing body of MCN literature consisted mostly of flawed studies contaminated with IPMNs. The goal of the present study is to clarify the clinical features, malignant potential, and long-term outcomes of MCNs defined by the presence of ovarian stroma in the largest single-institution, pathologyconfirmed series to date.
Methods Patients After obtaining approval from the Johns Hopkins Institutional Review Board, cases were identified from a prospectively managed pathology database. All potential MCNs were reviewed under the direct supervision of a single senior pathologist specializing in pancreatic pathology (RHH). Study inclusion was contingent upon the identification of ovarian stroma and confirmed cases were regraded according to the 2010 WHO guidelines [9]. A comparison cohort of resected IPMNs was also identified consisting of cases resected after the 1996 WHO classification through 2015. Reresections, cases with an unspecified grade of dysplasia, and IPMNs diagnosed secondary to an unrelated primary malignancy were all excluded. Clinical data for all study subjects were collected from the electronic medical record and original paper charts.
Analysis Statistical analyses were performed using the Stata 13 statistical software package (StataCorp. 2013. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP) and GraphPad Prism version 7.00 for Mac (GraphPad Software, La Jolla California USA, www.graphpad.com). All datasets were non-normally distributed based on the Shapiro-Wilk W test and nonparametric tests of significance were used for analyses. All analyses were performed as two-tailed tests with a significance threshold of a ¼ 0.05. Categorical data were expressed as relative frequencies (%) with total observations (n) and evaluated using Fisher's exact test or the chi-square test. Continuous data were summarized as medians with interquartile ranges (IQR) and compared using the Wilcoxon or Kruskal-Wallis rank tests depending on the number of comparison groups. For matched datasets, analogous evaluations were performed using either the Wilcoxon signed-rank test, McNemar test, or Stuart-Maxwell marginal homogeneity test. The last is a generalization of the McNemar test that allows for comparison of multilevel categorical variables in matched pairs.
Survival analysis Overall (OS) and disease-specific survival (DSS) were assessed by the Kaplan-Meier method with log-rank comparisons between cohorts and subgroups of interest. Survival was defined as the total time in years between resection and death or date of censoring. DSS includes only those deaths attributed to pancreatic cancer or its complications. Follow-up time and survival were calculated using the most recent confirmed medical encounter based on our records. Patients without confirmed deaths were not assumed to be alive at the time of publication and were censored at their most recent follow-up. Deaths were determined by chart review and through the Social Security Death Index (SSDI) with causes obtained from the medical record and death certificates. Results Demographics and clinicopathological features From 1984 through 2015, 142 pathology-confirmed MCNs were resected at the Johns Hopkins Hospital, accounting for <2% of all pancreatic resections during that timeframe. Of 142 MCNs, 101 (71.1%) had low-grade dysplasia (LGD), 15 (10.6%) intermediategrade dysplasia (IGD), 12 (8.5%) high-grade dysplasia (HGD), and 14 (9.9%) had an associated invasive cancer (ICA). The IPMN cohort consisted of 746 cases resected between 1996 and 2015. Pathology results included 103 (13.8%) cases with LGD, 247 (33.1%) with IGD, 154 (20.6%) with HGD, and 242 (32.4%) with an associated ICA. Detailed patient demographics and clinicopathological data are presented in Table 1. Males with an MCN were extremely uncommon in our series (n ¼ 5, 3.5%) and tended to be much older than their female counterparts [59.0 (56.0e61.0) years vs 47.0 (37.0e57.0) years, p ¼ 0.03]. Among female patients, 66 (48.2%) were premenopausal and 71 (51.8%) were peri- (n ¼ 6) or postmenopausal (n ¼ 65). Ten (14.9%) premenopausal women presented during or shortly after pregnancy and 16 (22.5%) peri/postmenopausal patients reported a history of estrogen-based HRT. No baseline demographics evaluated in this study were associated with the pathological finding of invasive disease. However, MCNs with an associated ICA were significantly larger on pathology compared to those without an associated ICA (p ¼ 0.0002). As demonstrated by Fig. 1, increasing tumor size was also associated with increasing grade of dysplasia (p ¼ 0.0001). Clinical features in IPMNs associated with the presence of ICA included advanced age (p ¼ 0.003), decreased body mass index (BMI) (p ¼ 0.001), history of heavy alcohol use (p ¼ 0.04), preoperative diabetes mellitus (p ¼ 0.003), and American Society of Anesthesia (ASA) class (p < 0.001). For the purpose of this study, “heavy alcohol use” was defined as a documented history of alcoholic pancreatitis, alcohol dependence, or reported consumption at or approaching daily use. Surgical resections for IPMNs with an associated ICA consisted of more TPs and fewer DPs compared to IPMNs without an associated ICA (p < 0.001). Invasive disease was also associated with greater lymph node yield (p < 0.0001) and larger tumor size (p < 0.001) on pathology. MCN presenting symptoms Patients were considered symptomatic if complaints were consistent with a pancreatic source defined as left upper quadrant (LUQ) or epigastric pain in the absence of another obvious and identifiable causes (e.g., cholelithiasis, diverticulitis, nephrolithiasis). Abdominal pain was the most common presenting complaint (n ¼ 93, 65.5%), but only 72 (50.7%) had pancreas-type
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
3
Table 1 Demographics and clinicopathological features of the study population. Baseline Demographics
MCN All
IPMN Invasive
Noninvasive
Pc
MCN vs IPMN
All
Invasive
Noninvasive
Pd
Pe
Total Subjects, (n)
142
14
128
746
242
504
Age, median (IQR)
47.5 (37.0 e57.0)
42.0 (36.8 e57.5)
48.0 (37.0 e57.0)
0.68b
69.0 (61.0 e75.0)
70.0 (62.0 e77.0)
68.0 (60.0 e74.0)
0.003b
<0.0001b
Female, % (n):
96.5% (137)
92.9% (13)
96.9% (124)
0.41a
49.6% (370)
44.6% (108)
52.0% (262)
0.06a
<0.001a
Race, % (n): Caucasian African American Asian Other
71.8% (102) 18.3% (26) 4.9% (7) 4.9% (7)
92.9% (13) 7.1% (1) 0.0% (0) 0.0% (0)
69.5% (89) 19.5% (25) 5.5% (7) 5.5% (7)
0.19a 0.31a 1.0a 1.0a
89.0% (664) 5.4% (40) 2.1% (16) 3.5% (26)
90.5% (219) 5.8% (14) 1.2% (3) 2.5% (6)
88.3% (445) 5.2% (26) 2.6% (13) 4.0% (20)
0.52a 0.73a 0.29a 0.63a
<0.001a <0.001a 0.07a 0.23a
BMI, median (IQR)
27.6 (23.6 e30.9)
29.1 (24.3 e30.6)
27.5 (23.1 e32.3)
0.71b
25.7 (22.8 e29.0)
24.6 (22.0 e27.2)
26.4 (23.1 e29.5)
0.001b
0.001b
Smoking History, % (n):
47.9% (68)
64.3% (9)
46.1% (59)
0.27a
56.9% (413)
56.2% (136)
55.0% (277)
0.69a
0.11a
Heavy Alcohol Use, % (n):
12.7% (18)
14.3% (2)
12.5% (16)
1.0
a
19.5% (126)
22.3% (54)
14.3% (72)
0.04
a
a
22.3% (166)
28.9% (70)
19.0% (96)
0.003
0.09a a
<0.001a
Diabetes Mellitus, % (n):
7.7% (11)
0.0% (0)
8.6% (11)
0.60
ASA Class, % (n): I II III IV
3.8% (4) 61.9% (65) 34.3% (36) 0% (0)
7.1% (1) 35.7% (5) 35.7% (5) 0.0% (0)
2.3% (3) 46.9% (60) 24.2% (31) 0.0% (0)
0.21a
0.2% (1) 32.6% (210) 65.5% (422) 1.7% (11)
0.0% (0) 20.2% (49) 62.4% (151) 3.3% (8)
0.2% (1) 31.9% (161) 53.8% (271) 0.6% (3)
<0.001a
<0.001a
Extra-Pancreatic Malignancy, Any Breast (% females) Prostate (% males)
% (n): 12.0% (17) 6.6% (9) 20% (1)
21.4% (3) 15.4% (2) 0% (0)
10.2% (13) 5.6% (7) 25% (1)
0.20a
21.0% (157) 7.6% (28) 10.4% (39)
19.0% (46) 1.9% (2) 10.4% (14)
22.0% (111) 9.9% (26) 10.3% (25)
0.50a
0.01a
78.6% (11) 14.3% (2) 7.1% (1) 0.0% (0) 0.0% (0)
91.4% (117) 4.7% (6) 0.8% (1) 0.8% (1) 2.3% (3)
0.15a
24.3% (181) 65.1% (486) 8.7% (65) 1.3% (10) 0.5% (4)
17.0% (41) 64.3% (155) 18.7% (45) 0.0% (0) 0.0% (0)
27.8% (140) 65.4% (329) 4.0% (20) 2.0% (10) 0.8% (4)
<0.001a
<0.001a
Surgical Resections & Pathology Resection Type, % (n): DP PD TP CP Enucleation
90.1% (128) 5.6% (8) 1.4% (2) 0.7% (1) 2.1% (3)
Tumor Size (cm), median (IQR) 4.2 (2.5e7.0)
10.0 (4.9e14.0) 4.0 (2.5e6.4)
0.0002b 2.5 (1.5e3.5)
3.0 (1.9e5.2)
2.3 (1.5e3.0)
<0.0001b <0.0001b
Nodal Status: Total LN, Median (IQR) Positive LN, % (n)
13 (6e21.3) 0.7% (1)
16 (10.5e25) 7.1% (1)
12.5 (6e21) 0.0% (0)
0.23b
17 (13e23) 16.5% (123)
21 (14e28) 50.8% (123)
15 (12e21) 0.0% (0)
<0.0001b <0.0001b 0.001a
Invasion, % (n): Lymphovascular Perineural
0.7% (1) 1.4% (2)
7.1% (1) 14.3% (2)
0.0% (0) 0.0% (0)
10.9% (81) 20.0% (149)
33.5% (81) 61.6% (149)
0.0% (0) 0.0% (0)
0.004a <0.001a
Significant p-values (<0.05) are in bold. Note. BMI, body mass index; ASA, America Society of Anesthesiologists; IQR, interquartile range; MPD, main pancreatic duct; DP, distal pancreatectomy; PD, pancreaticoduodenectomy; TP, total pancreatectomy; CP, central pancreatectomy; LN, lymph node; MCN, mucinous cystic neoplasm; P, p-value; IPMN, intraductal papillary mucinous neoplasm. a Fisher's Exact; b Mann-Whitney U. c Comparison of invasive vs noninvasive MCN. d Comparison of invasive vs noninvasive IPMN. e Comparison of total MCN vs total IPMN.
pain. Additional complaints included pain radiating to the back (n ¼ 30, 21.1%), nausea (n ¼ 33, 23.2%), post-prandial symptoms such as worsening pain and early satiety (n ¼ 26, 18.3%), vague gastrointestinal complaints including crampiness, gas, or bloating (n ¼ 24, 16.9%), weight loss (n ¼ 20, 14.1%), and diarrhea (n ¼ 14, 9.9%). Additionally, 28 (19.7%) were diagnosed with acute pancreatitis and 21 (14.8%) had a palpable abdominal mass on physical exam. The presence of an associated ICA was associated with a palpable mass on exam (p < 0.001) and postprandial symptoms (p ¼ 0.005). Tumor size was associated with a palpable mass and post-prandial symptoms as well (p < 0.0001 and p ¼ 0.002,
respectively), along with nausea (p ¼ 0.014 and diarrhea (p ¼ 0.031). Abdominal pain demonstrated no significant associations.
Diagnostic imaging analysis Imaging reports from preoperative computed tomography (CT) were reviewed for 139 MCN patients and 457 IPMN patients (Table 2). For patients with an MCN, CT findings associated with ICA included increasing tumor size (p ¼ 0.0003), mural nodules (p ¼ 0.004), internal septations (p ¼ 0.03), and solid or solid-cystic features in place of the typical macrocystic features (p ¼ 0.004).
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
4
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
Fig. 1. Comparison of MCN and IPMN tumor size by grade of dysplasia. Tumor size was obtained from computed tomography (CT) scans (left) and pathology (right). Points correspond to median (IQR) tumor sizes, which are recorded beneath the corresponding grade of dysplasia. P-values were obtained from Mann-Whitney U comparison of MCN and IPMN tumor size at each grade of dysplasia. LGD ¼ low-grade dysplasia, IGD ¼ intermediate-grade dysplasia, HGD ¼ high-grade dysplasia, and ICA ¼ invasive carcinoma.
Table 2 CT imaging features. CT Imaging Features
MCN
IPMN
MCN vs IPMN
All
Invasive
Noninvasive
P
All
Invasive
Noninvasive
P
Pe
Size (cm): Median (IQR)
4.0 (3.0e6.6)
10.0 (5.0e16.1)
3.8 (3.0e6.1)
0.0003b
3.0 (2.0e4.0)
3.3 (1.9e4.7)
2.9 (2.0e3.8)
0.047b
<0.0001b
Location, % (n): Proximal Distal Proximal & Distal
7.9% (11) 92.1% (128) 0% (0)
21.4% (3) 78.6% (11) 0% (0)
6.4% (8) 93.6% (117) 0% (0)
0.08a 0.08a e
67.6% (309) 24.9% (114) 7% (32)
68.2% (88) 24.8% (32) 7% (9)
67.4% (221) 25% (82) 7% (23)
0.91a 1.0a 1.0a
<0.001a <0.001a <0.001a
100% (139) 94.2% (131) 9.4% (13)
100% (14) 71.4% (10) 35.7% (5)
100% (125) 96.8% (121) 6.4% (8)
e 0.004a 0.004a
64.8% (296) 70.2% (321) 22.1% (101)
61.2% (79) 51.2% (66) 41.1% (53)
57% (187) 77.7% (255) 14.6% (48)
0.83a <0.001a <0.001a
<0.001a <0.001a 0.001a
38.1% (53) 12.9% (18) 2.9% (4) 0% (0)
64.3% (9) 0% (0) 14.3% (2) 0% (0)
35.2% (44) 14.4% (18) 1.6% (2) 0% (0)
0.04a 0.22a 0.05a e
37.2% (170) 5.5% (25) 46.4% (212) 68.7% (314)
32.6% (42) 6.2% (8) 69.8% (90) 62% (80)
39% (128) 5.2% (17) 37.5% (123) 71.6% (235)
0.24a 0.65a <0.001a 0.06a
0.84a 0.005a <0.001a <0.001a
Features, % (n): Solitary Lesion Macrocystic Mural Nodule/Solid Component Size (mm) Septations Calcifications Dilated MPD MPD Communication
c
d
Significant p-values (<0.05) are in bold. Note. CT, computed tomography; IQR, interquartile range; MPD, main pancreatic duct; MCN, mucinous cystic neoplasm; P, p-value; IPMN, intraductal papillary mucinous neoplasm. a Fisher's Exact; b Mann-Whitney U. c Comparison of invasive vs noninvasive MCN. d Comparison of invasive vs noninvasive IPMN. e Comparison of total MCN vs total IPMN.
CT evidence of splenic vein thrombosis was also strongly correlated with MCN-associated ICA. After excluding proximal MCNs, there were 98 cases (72 LGD, 10 IGD, 8 HGD, 8 ICA) with CT reports commenting on splenic vein (SV) patency. Thrombosis was observed in all 8 (100%) MCNs with an associated ICA compared with only 10 (11.1%) out of 90 MCNs without ICA (p<0.001). For those with an IPMN, CT findings associated with ICA included increasing tumor size (p ¼ 0.047), mural nodules (p < 0.001), absence of macrocystic features (p < 0.001), and main pancreatic duct (MPD) dilation (p < 0.001).
Disease-specific survival Detailed DSS data are presented in Table 3 with corresponding Kaplan-Meier curves in Fig. 2. There were only 2 disease-specific deaths in the entire MCN cohort; both occurred in the setting of an associated ICA, 1 of which had positive lymph nodes. Absence of an associated ICA demonstrated DSS of 100% at all time points, while presence of an associated ICA (all cases) had DSS of 90.9%, 79.6%, and 79.6% at 1, 5, and 10 years, respectively. Considering only N0 invasive disease, DSS was 100%, 87.5%, and 87.5%, respectively. For the IPMN cohort, disease-specific deaths were confirmed in 17 (3.4%) patients without an associated ICA, 124 (51.2%) patients
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
5
Table 3 Summary results of Kaplan-Meier survival analyses. Full Cohort
Invasive (N0 Only)
Noninvasive
MCN
IPMN
MCN
Invasive (All) IPMN
MCN
IPMN
MCN
IPMN
Overall Survival At risk, (n) Median (IQR), y 95% CI Deaths, % (n) 1-year, % 5-year, % 10-year, % Log-Rank (P)
140 e e 7.1% (10) 98.9% 88.1% 80.9% <0.0001
746 7.8 (2.9e13.6) 6.7e8.9 32.8% (245) 91.0% 62.3% 37.5%
14 e e 14.3% (2) 90.9% 79.6% 79.6% 0.0003
242 2.4 (1.2e6.4) 2.0e3.2 64.0% (155) 79.2% 31.2% 12.5%
12 e e 8.3% (1) 100% 87.5% 87.5% 0.002
118 5.4 (2.0e8.8) 3.8e7.7 51.7% (61) 86.9% 51.1% 22.0%
126 e e 6.3% (8) 100% 89.6% 80.4% 0.0008
504 10.9 (6.8e15.4) 9.8e13.8 17.8% (90) 97.4% 81.6% 54.8%
Disease-Specific At risk, (n) Median (IQR), y 95% CI Deaths, % (n) 1-year, % 5-year, % 10-year, % Log-Rank (P)
140 e e 1.4% (2) 98.9% 96.8% 96.8% <0.0001
746 e e 18.9% (141) 92.7% 72.9% 63.1%
14 e e 14.3% (2) 90.9% 79.6% 79.6% 0.01
242 2.7 (1.3-.) 2.0e3.8 51.2% (124) 80.3% 36.0% 27.2%
12 e e 8.3% (1) 100% 87.5% 87.5% 0.058
118 7.8 (2.0-.) 2.2e4.4 33.9% (40) 88.7% 57.7% 44.3%
126 e e 0% (0) 100% 100% 100% 0.052
504 e e 3.4% (17) 99.5% 95.9% 86.2%
Significant p-values (<0.05) are in bold. Note. IQR, interquartile range; CI, confidence interval; P, p-value; MCN, mucinous cystic neoplasm; IPMN, intraductal papillary mucinous neoplasm.
with an associated ICA, and 40 (33.9%) with N0 ICA. DSS at 1, 5, and 10 years was 99.5%, 95.9%, and 86.2%, respectively for noninvasive IPMNs, 80.3%, 36.0%, and 36.0%, respectively for those with an associated ICA (all cases), and 88.7%, 57.7%, and 44.3% respectively for N0 invasive disease. The log-rank test demonstrated significantly better DSS for patients with MCNs compared to those with IPMNs overall (p < 0.0001) and in the setting of an associated ICA (p ¼ 0.02), while N0 disease (p ¼ 0.58) and noninvasive disease (p ¼ 0.052) approached significance. Overall survival Detailed OS data, including subanalyses by invasiveness, are presented in Table 3 with corresponding Kaplan-Meier curves in Fig. 2. There were 10 (7.0%) deaths overall in the MCN cohort. Eight occurred in patients with noninvasive disease from causes unrelated to their pancreatic neoplasms, while the other two patients died from MCN-associated ICA. OS at 1, 5, and 10 years was 98.9%, 88.1%, and 80.9%, respectively. There were 245 (32.8%) total deaths recorded in the IPMN cohort, 90 of which occurred in patients with noninvasive disease and 155 in patients with an IPMN with associated ICA. Median OS for the entire IPMN cohort was 7.8 (2.9e13.6) years with 1-, 5-, and 10-year OS of 91.0%, 62.3%, and 37.5%, respectively. Log-rank comparison confirmed significantly greater OS for patients with an MCN compared to those with an IPMN (p < 0.0001). OS in the subsets of noninvasive and invasive disease also demonstrated significantly higher survival for MCNs compared to IPMNs over the 1-, 5-, and 10-year intervals (Table 3). Discussion This study was conducted at a high-volume center for pancreatic diseases and comprises the largest single-institution series of pathology-confirmed MCNs to date. Based on our experience, MCNs have a highly stereotypical clinical profile presenting almost exclusively in middle-aged women as solitary, macrocystic lesions in the body and tail of the pancreas. By comparison, IPMNs usually present as proximal tumors in older patients with a nearly equal sex distribution. Survival analysis demonstrated a much better prognosis for
patients with an MCN compared to those with an IPMN due to the infrequent progression of MCNs to ICA and better survival once ICA develops. The overwhelming majority of MCNs presented with noninvasive disease, all of which were cured by surgical resection. However, a more interesting observation is that patients with MCNassociated ICA also had a favorable prognosis, particularly compared to the 51.2% disease-specific mortality rate of patients with IPMN-associated ICA. Only 2 (14.3%) of the 14 patients with an MCN-associated ICA died, and all 12 survivors remained alive without evidence of recurrence at the time of this study. Included among these were several examples of long-term survival exceeding 10 years (n ¼ 3) and 20 years (n ¼ 2) with an overall median follow-up time of 42.8 (6.4e166.7) months. Survival for patients with a surgically resected MCN or IPMN was closely correlated with pathology results, particularly lymph node involvement, lymphovascular invasion, and perineural invasion. These are high-risk pathologic features that signal the presence of more advanced, likely micrometastatic disease associated with a high rate of recurrence and poor prognosis [10,11]. Though very common among IPMN-associated ICA, high-risk pathologic features were rarely identified in MCN-associated ICA, which likely contributes to their favorable prognosis. Comparable results were reported by Crippa et al. in 19 MCNs with associated ICA [12]. Similarly, a recent study from this institution suggested that the presence of minimally-invasive disease limited to the ovarian stroma is not a poor prognosticator in MCNs [13]. Recent studies have suggested the possibility of nonoperative management for small (<3e4 cm) MCNs without high-risk features, particularly in borderline surgical candidates [4,12]. In our series, the smallest MCN-associated ICA measured 4.0 cm on CT. There were 41 (28.9%) MCNs measuring <4 cm without mural nodules, including 40 with LGD and a single MCN with IGD. These results suggest that imaging criteria can effectively identify lowgrade lesions and close imaging surveillance may be a reasonable management option for small (<4 cm) MCN without concerning imaging features. Furthermore, since only 2 (14.3%) of the 14 patients with MCN-associated ICA died from their disease, even those progressing during surveillance may still have a very good prognosis. However, since MCNs generally present in otherwise young and healthy patients, “lifelong surveillance” could entail several
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
6
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
Fig. 2. Kaplan-Meier survival curves depicting survival among MCN (dashed) and IPMN (solid). I.) Overall and disease-specific survival in the full MCN and IPMN cohorts. Log-rank comparison demonstrated significantly better overall (A) and disease-specific (B) survival for MCN compared to IPMN (p < 0.0001 and p < 0.0001, respectively). II.) Disease-specific survival in MCN and IPMN with (C) and without (D) an associated ICA. Log-rank comparison demonstrated significantly better disease-specific survival for MCN-associated ICA versus IPMN-associated ICA overall (p ¼ 0.02) and approached significance when comparing only N0 disease (p ¼ 0.058). Log-rank comparison of MCN and IPMN without an associated ICA also approached significance (p ¼ 0.052).
decades of imaging follow-up and most patients present as excellent surgical candidates (65.7% ASA I/II, Table 1). Therefore, when considering nonoperative management, surgeons must weigh the associated factors of cost-effectiveness, radiation exposure, compliance with very long-term follow-up, and potential for disease progression and death against the fact that our series demonstrated resection to be curative for all cases of non-invasive disease with no operative mortality. The overwhelming likelihood is that most patients do not require surgical management, but there is currently no long-term data available to suggest which patients are at risk for progression to ICA. Therefore, larger series and better long-term data are required to identify those patients who would benefit most from surgery. The vast majority of MCNs arise in women and some studies have linked them to pregnancy and hormone-replacement therapy
(HRT) [14e19]. As a result, some have speculated that MCN may be hormonally influenced. In our series, we did not observe any statistically significant associations with pregnancy or HRT. However, we made the interesting observation that premenopausal women accounted 9/13 (69.2%) female MCN-associated ICAs. If validated in other studies, this trend would support the notion of hormonal influence on tumor progression and suggest that premenopausal patients are at greatest risk for cancer. Furthermore, a reduction in risk following menopause could have significant implications for current management recommendations. As it currently stands, this is an isolated, nonsignificant finding and requires additional study in a much larger series. As with any retrospective study, our data were limited to data available in the medical record. The fact that patients often originated as referrals from other institutions meant that records were
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003
J.F. Griffin et al. / Pancreatology xxx (2017) 1e7
often limited for data unrelated to their primary pancreatic disease. These patients usually transitioned care closer to home following surgery, which also decreased the availability of long-term data and follow-up. This is particularly true for MCN patients, who often did not require long-term surgical follow-up. One final limitation of the study was the rarity of MCNs, which greatly limits the size of a single institution series.
[6] [7]
[8]
Conclusion [9]
This is the largest single-institution series of MCNs reported to date with 142 cases confirmed by the presence of ovarian stroma on pathology review. Based on our results, MCNs are rare tumors with a consistent presentation and infrequent progression to cancer. For MCNs without an associated cancer, complete surgical resection was curative in 100% of cases. For those with MCN-associated ICA, prognosis was still excellent with a 14.3% mortality rate and 10-year survival approaching 80%.
[10]
[11]
[12]
Acknowledgements Supported by grants from the National Cancer Institute of the National Institutes of Health (T32CA126607 and CA62824), the Michael Rolfe Foundation, and the Sol Goldman Pancreatic Cancer Research Center. The authors have no conflicts of interest to disclose.
[13]
[14]
[15]
References [16] [1] Compagno J, Oertel JE. Mucinous cystic neoplasms of the pancreas with overt and latent malignancy (cystadenocarcinoma and cystadenoma). A clinicopathologic study of 41 cases. Am J Clin Pathol 1978;69:573e80. [2] Ohashi K, Murakami Y, Takeoshi T, Ohta HOI. Four cases of mucin producing cancer of the pancreas on specific findings of the papilla of Vater. Prog Dig Endosc 1982;20:348e51. [3] Itai Y, Ohhashi K, Nagai H, Murakami Y, Kokubo T, Makita K, et al. “Ductectatic” mucinous cystadenoma and cystadenocarcinoma of the pancreas. Radiology 1986;161:697e700. http://dx.doi.org/10.1148/radiology.161.3.3786719. [4] Reddy RP, Smyrk TC, Zapiach, M, Levy MJ, Pearson RK, Clain JE, et al. Pancreatic mucinous cystic neoplasm defined by ovarian stroma: demographics, clinical features, and prevalence of cancer. Clin Gastroenterol Hepatol 2004;2: 1026e31. http://dx.doi.org/10.1016/S1542-3565(04)00450-1. [5] de Wilde RF, Hruban RH, Maitra A, Offerhaus GJA. Reporting precursors to
[17]
[18]
[19]
7
invasive pancreatic cancer: pancreatic intraepithelial neoplasia, intraductal neoplasms and mucinous cystic neoplasm. Diagn Histopathol 2012;18:17e30. http://dx.doi.org/10.1016/j.mpdhp.2011.10.012. Lennon AM, Wolfgang C. Cystic neoplasms of the pancreas. J Gastrointest Surg 2013;17:645e53. http://dx.doi.org/10.1007/s11605-012-2072-6. Kloppel G, Solicia E, Longnecker D, Capella C, Sobin L. Histologic typing of tumors of the exocrine pancreas: World Health organization international histologic classification of tumors. second ed. Berlin: Springer-Verlag; 1996. Tanaka M, Chari S, Adsay V, Fernandez-del Castillo C, Falconi M, Shimizu M, et al. International consensus guidelines for management of intraductal papillary mucinous neoplasms and mucinous cystic neoplasms of the pancreas. Pancreatology 2006;6:17e32. http://dx.doi.org/10.1159/000090023. Zamboni G, Fukushima N, Hruban RH. Mucinous cystic neoplasms of the pancreas. In: Bosman FE, Carneiro F, Hruban H, Theise N, editors. WHO Classif. Tumors Dig. Syst. fourth ed. Lyon: IARC; 2010. p. 300e3. Winter JM, Cameron JL, Campbell KA, Arnold MA, Chang DC, Coleman J, et al. 1423 pancreaticoduodenectomies for pancreatic cancer: a single-institution experience. J Gastrointest Surg 2006;10:1199e210. http://dx.doi.org/10.1016/ j.gassur.2006.08.018. Nathan H, Wolfgang CL, Edil BH, Choti MA, Herman JM, Schulick RD, et al. Perioperative mortality and long-term survival after total pancreatectomy for pancreatic adenocarcinoma: a population-based perspective. J Surg Oncol 2009;99:87e92. http://dx.doi.org/10.1002/jso.21189. Crippa S, Salvia R, Warshaw AL, Domínguez I, Bassi C, Falconi M, et al. Mucinous cystic neoplasm of the pancreas is not an aggressive entity: lessons from 163 resected patients. Ann Surg 2008;247:571e9. http://dx.doi.org/ 10.1097/SLA.0b013e31811f4449. Lewis GH, Wang H, Bellizzi AM, Klein AP, Askin FB, Schwartz LE, et al. Prognosis of minimally invasive carcinoma arising in mucinous cystic neoplasms of the pancreas. Am J Surg Pathol 2013;37:601e5. http://dx.doi.org/10.1097/ PAS.0b013e318273f3b0. Kosumi K, Takamori H, Hashimoto D, Tanaka H, Abe S, Nakahara O, et al. Mucinous cystic neoplasm of the pancreas activated during pregnancy. Surg Case Rep 2015;1:13. http://dx.doi.org/10.1186/s40792-014-0012-2. Tica AA, Tica OS, Saftoiu A, Camen D, Tica VI. Large pancreatic mucinous cystic neoplasm during pregnancy: what should be done? Gynecol Obstet Invest 2013;75:132e8. http://dx.doi.org/10.1159/000346176. Tsuda H, Kotani T, Sumigama S, Mano Y, Shimoyama Y, Kikkawa F. Mucinous cystic neoplasm of the pancreas with severe dysplasia during pregnancy: case report and review of the literature. Taiwan J Obstet Gynecol 2012;51:635e8. http://dx.doi.org/10.1016/j.tjog.2012.09.022. Kosumi K, Takamori H, Hashimoto D, Tanaka H, Abe S, Nakahara O, et al. Mucinous cystic neoplasm of the pancreas activated during pregnancy. Surg Case Rep 2015;1:13. http://dx.doi.org/10.1186/s40792-014-0012-2. Tanaka S, Kawamura T, Nakamura N, Teramoto K, Arii S. Mucinous cystadenocarcinoma of the pancreas developing during hormone replacement therapy. Dig Dis Sci 2007;52:1326e8. http://dx.doi.org/10.1007/s10620-0069482-9. Foster D, Shaikh MF, Gleeson E, Babcock BD, Lin J, Ownbey RT, et al. Pancreatic mucinous cystic neoplasm in a transgender patient. World J Surg Oncol 2015;13:205. http://dx.doi.org/10.1186/s12957-015-0620-8.
Please cite this article in press as: Griffin JF, et al., Patients with a resected pancreatic mucinous cystic neoplasm have a better prognosis than patients with an intraductal papillary mucinous neoplasm: A large single institution series, Pancreatology (2017), http://dx.doi.org/10.1016/ j.pan.2017.04.003