Assessment of clonality of multisegmental main duct intraductal papillary mucinous neoplasms of the pancreas based on GNAS mutation analysis Koji Tamura, MD,a Takao Ohtsuka, MD, PhD,a Taketo Matsunaga, MD,a Hideyo Kimura, MD,a Yusuke Watanabe, MD,a Noboru Ideno, MD, PhD,a Teppei Aso, MD, PhD,a Tetsuyuki Miyazaki, MD,b Kenoki Ohuchida, MD, PhD,a Shunichi Takahata, MD, PhD,a Tetsuhide Ito, MD, PhD,c Yasuhiro Ushijima, MD, PhD,d Yoshinao Oda, MD, PhD,b Kazuhiro Mizumoto, MD, PhD,a and Masao Tanaka, MD, PhD, FACS,a Fukuoka, Japan
Background. Main duct intraductal papillary mucinous neoplasms (MD-IPMNs) may occur in 1 or multiple segments of the pancreatic duct. Unlike multifocal branch duct (BD)-IPMNs, the clonality of multisegmental MD-IPMNs remains unclear. GNAS mutations are common and specific for IPMNs, and mutational assessment might be useful to determine the clonality of IPMNs as well as to detect high-risk IPMN with distinct ductal adenocarcinoma (pancreatic ductal adenocarcinoma [PDAC]). Our aim was to clarify clonality using GNAS status in multisegmental MD-IPMNs. Methods. We retrospectively reviewed the medical records of 70 patients with MD-IPMN. Histologic subtypes and KRAS/GNAS mutations were investigated, and the clonal relationships among multisegmental MD-IPMNs were assessed. Mutational analysis was performed using high-resolution melting analysis and subsequent Sanger/pyrosequencing. Results. Thirteen patients had multiple synchronous and/or metachronous lesions. Seven of these 13 patients had multiple MD-IPMNs; 3 had multiple MD-IPMNs and distinct BD-IPMNs; 1 had multiple MD-IPMNs and a distinct PDAC; 1 had a solitary MD-IPMN, BD-IPMN, and PDAC; and 1 had a solitary MD-IPMN and PDAC. KRAS/GNAS mutations were consistent in 10 of 11 multisegmental MD-IPMNs, whereas MD-IPMNs, BD-IPMNs, and PDACs tended to show different mutational patterns. The frequency of malignant IPMNs was significantly higher in the multisegment cohort; malignant IPMNs constituted 90% (9/10) of the multiple cohort and 56% (32/57) of the solitary cohort (P = .04). Mutant GNAS was more frequently observed in the intestinal subtype (94%) than the others. Conclusion. MD-IPMNs can be characterized by monoclonal skip progression. Close attention should be paid to the possible presence of skip areas during or after partial pancreatectomy. (Surgery 2015;157:277-84.) From the Departments of Surgery and Oncology,a Anatomic Pathology,b Medicine and Bioregulatory Science,c and Clinical Radiology,d Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan
Conflicts of interest and source of funding: The authors declare no conflicts of interest. This study was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Numbers 24390318, 24390319, 23390327, 25293285, 25670586, 25670585, 25670584, 25670582, and 24592030. Accepted for publication September 10, 2014. Reprint requests: Masao Tanaka, MD, PhD, FACS, Department of Surgery and Oncology, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 8128582, Japan. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2014.09.013
SEVERAL STUDIES have reported that intraductal papillary mucinous neoplasms of the pancreas (IPMNs) are characterized by multifocality, the prevalence of which varies from 20 to 41%.1,2 Most previous studies assessed both main duct (MD-IPMNs) and branch duct IPMNs (BD-IPMNs), and the majority of multifocal IPMNs were of the BD type. A field carcinogenesis effect seems to play an important role in some cases of IPMN.3 Z’graggen et al4 reported that KRAS mutations were often detected in the histologically normal epithelium of some IPMN patients, which seems to demonstrate field malignant transformation in the entire pancreas SURGERY 277
278 Tamura et al
of IPMN patients. On the basis of such genetic alterations, several studies examining multifocal lesions of IPMN in the same pancreas showed different patterns of KRAS mutations or loss of heterozygosity, suggesting that multifocal IPMNs may originate from independent precursor lesions.3,5 Conversely, we previously suggested that MDIPMNs might demonstrate monoclonal progression because the mutational patterns of recurrent lesions in the remnant pancreas were consistent with those of the initial lesions in 4 MD-IPMN patients, although the surgical margin was negative for neoplastic cells.6 This finding seems to differ from the polyclonality of BD-IPMNs. MD-IPMN may occur in 1 or multiple segments of the main pancreatic duct; however, the clonality of the genetic alterations of multisegmental MD-IPMNs remains unclear. Several epithelial malignancies reportedly arise through an adenoma–carcinoma sequence.7 IPMNs are well known to show a wide spectrum of histologic differentiation from adenomas and borderline neoplasms to carcinoma, and are thus considered to be a good model of the adenoma– carcinoma sequence. GNAS is a gene encoding the guanine nucleotide-binding protein a subunit. Recent reports have stated that GNAS mutations are common and specific for IPMNs, whereas other pancreatic neoplasms, including pancreatic ductal adenocarcinoma (PDAC), rarely have GNAS mutations.8,9 Wu et al8 found that 47.8% of low-grade IPMNs had GNAS mutations, indicating that GNAS mutation might be associated with tumor initiation as a founder mutation of IPMNs. In this study, the mutation findings and histologic subtypes among distinct synchronous/metachronous multiple MD segments and the GNAS mutational status among MD-IPMNs were comprehensively analyzed to clarify the clonal evolution and genetic alterations among MD-IPMNs. PATIENTS AND METHODS The present study was approved by the Ethics Committee of Kyushu University and conducted according to the Ethical Guidelines for Human Genome/Gene Research enacted by the Japanese Government and the Declaration of Helsinki. We reviewed retrospectively the medical records of 70 patients who had been histologically diagnosed with MD-IPMN after resection at the Department of Surgery and Oncology, Kyushu University Hospital between 1987 and 2013. The data of the initial 56 patients were described in our previous study for different purposes.6 MD-IPMN was
Surgery February 2015 defined as >5-mm segmental or diffuse dilation of the main pancreatic duct (MPD) without >5mm dilation of the BD.2 Mixed-type IPMN was included as an MD-IPMN in this study, as previously reported.6 In accordance with the 2010 World Health Organization criteria, IPMN components were classified into 4 different histologic grades: Lowgrade dysplasia, intermediate-grade dysplasia (IGD), high-grade dysplasia (HGD), and invasive carcinoma (Inv-CA).10 IPMNs were also classified into 4 histologic subtypes: Gastric, intestinal, pancreatobiliary (PB), and oncocytic. When multiple IPMNs were simultaneously present in the same pancreas, the histologic grade and subtype were determined according to of the IPMN with the highest degree of dysplasia. All patients were classified into 1 of 2 cohorts according to the pathologic mapping of the whole specimen after resection: Solitary and multiple cohorts. A solitary MD-IPMN was defined as 1 continuous tumor, and a multisegmental MDIPMN was defined as a tumor affecting $2 distinct segments of the MPD separated from each other by an uninvolved segment of normal MPD. We confirmed that the MPD epithelium between the 2 distinct lesions had no papillary dysplastic changes. Additionally, there were 2 types of multiple lesions: Synchronous and metachronous tumors. Synchronous lesions were defined as multiple lesions that were resected during the same operation, and metachronous lesions were defined as distinct lesions that had metachronously developed in the remnant pancreas after partial pancreatectomy with negative surgical margins. KRAS and GNAS mutations and histologic subtypes of each lesion were investigated to assess the clonality of multisegmental MD-IPMNs. For the purpose of mutational analysis, appropriate tissue blocks were obtained from formalinfixed, paraffin-embedded tissue samples. Tumor tissues were dissected macroscopically if the neoplastic cellularity was $70%. They were microdissected with a sterile needle under an optical microscope (BX41 TF; Olympus, Tokyo, Japan) in cases of low neoplastic cellularity or for the normal pancreatic duct epithelium. Genomic DNA was subsequently extracted using the QIAamp DNA Formalin-Fixed, Paraffin-Embedded Tissue Kit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). The mutational status of exon 2 of KRAS and exon 8 of GNAS was investigated with high-resolution melting curve analysis followed by Sanger sequencing and/or pyrosequencing. Briefly, polymerase chain
Surgery Volume 157, Number 2
reactions for high-resolution melting curve analysis were carried out in a final volume of 10 mL, composed of 5 mL of SsoFast EvaGreen Supermix (Bio-Rad Laboratories, Hercules, CA); 3.2 mL of diethylpyrocarbonate-treated, pyrogen-free, DNase/RNase-free water (Invitrogen, Carlsbad, CA); 1 mL (10 ng/mL) of template DNA; and 0.4 mL (10 pmol/mL) of each forward and reverse primer pair. All samples were tested in triplicate. Polymerase chain reaction products that were indicated to have mutations by highresolution melting curve analysis were subjected to Sanger sequencing to confirm the single-nucleotide polymorphism hot spot as previously described.6 If a mutation was not detected by Sanger sequencing, additional pyrosequencing with the PyroMark Q24 (Qiagen) using PyroMark Gold reagents (Qiagen) was performed according to the manufacturer’s instructions. When single nucleotide polymorphisms were not detected by pyrosequencing, the allele was defined as wild type. The primers used in this study are listed in the Supplementary Table. Statistical analyses were completed by JMP statistical software (version 9.0.2; SAS Institute, Cary, NC). The Chi-square test or Fisher’s exact test was used to evaluate differences in categorical data between 2 cohorts, and the Mann–Whitney U test was used for continuous data. The disease-specific survival rate was analyzed using the Kaplan–Meier method with log-rank univariate comparisons. RESULTS Clinicopathologic characteristics. In total, 70 resected MD-IPMNs including 13 mixed IPMNs were reviewed. The histologic grades of these 70 IPMNs at the time of the initial operation were lowgrade dysplasia in 20 patients (29%), IGD in 7 (10%), HGD in 19 (27%), Inv-CA in 23 (33%; including 7 T1a, formerly called minimally invasive, carcinomas), and unknown in 1 (HGD or Inv-CA). The histologic subtypes were gastric type in 32 patients (47%), intestinal in 24 (35%), PB in 9 (13%), oncocytic in 3 (4%), and unknown in 2. Three PDACs (2 synchronous and 1 metachronous) were simultaneously present with MD-IPMNs, as previously reported.5 Seven patients developed metachronous MD-IPMNs in the remnant pancreas after partial pancreatectomy, 4 of which were free from neoplastic cells at the pancreatic margin and 3 of which were positive for atypical cells or had unknown details. There was no difference in either the cumulative disease-specific survival rate (P = .44; Supplementary Figure) or the rate of recurrence in the remnant pancreas (P = .96; Supplementary Figure) between patients with (n = 22) and without (n = 36) atypical cells at the resected margin.
Tamura et al 279
Analyses of KRAS and GNAS mutations and histologic subtype of multiple lesions in MD-IPMNs. Thirteen patients in this series had multiple synchronous (n = 9) and/or metachronous (n = 6) neoplastic lesions (Table I). Seven of these 13 patients had multiple-segment MD-IPMNs (patients 1–3, 5, and 8–10); 3 had multiple MD-IPMNs and a distinct BD-IPMN; 1 had multiple MD-IPMNs and a distinct PDAC; 1 had a solitary MD-IPMN, BD-IPMN, and PDAC; and 1 had a solitary MD-IPMN and distinct PDAC (Fig 1). In 10 of the 11 patients with multiple MD-IPMNs (excluding patient 7), the mutational patterns of the multiple segments of the MD-IPMNs were consistent with one another in the same pancreas (Table II; Fig 2). In patients 1 and 8, the histologic subtype of the main lesion differed from that of a sublesion (one was gastric and the other was PB type), although they seemed to be histologically similar. MD-IPMN and BD-IPMN lesions tended to show different mutational patterns and/or histologic subtypes, indicating a polyclonal relationship, as observed in patients 4, 7, and 11. There were also different mutational patterns between MD-IPMNs and concomitant PDACs. Additionally, when the histologic grade distribution was compared between multiple (n = 10) and solitary (n = 57) IPMN cohorts (excluding 3 patients with concomitant PDACs), the frequency of malignant IPMNs (HGD/Inv-CA) was significantly higher in the cohort with multiple segment MD-IPMNs; malignant IPMNs constituted 90% (9/10) of the multiple cohort and 56% (32/57) of the solitary cohort (P = .04). Effects of GNAS mutation on clinicopathologic features. A comprehensive GNAS mutational analysis of 53 MD-IPMNs resected between 2000 and 2013 was performed to investigate the significance of GNAS mutation in the biological behavior of MD-IPMN. Seventeen samples were excluded for such reasons as old formalin-fixed, paraffinembedded samples (obtained before 1999), inadequate DNA quality, insufficient amplification of the polymerase chain reaction product, and lack of adequate tissue samples. Thirty-six of 53 MD-IPMNs (68%) harbored a GNAS codon 201 point mutation, and 17 (32%) were determined to have wild-type GNAS. Clinical characteristics, such as patient age, symptoms, MPD diameter, presence of mural nodules, serum levels of carcinoembryonic antigen and carbohydrate antigen 19-9 (normal or high), and histologic grade and subtype, were compared between the GNAS wild-type and mutant cohorts. There was a significant difference in the subtype distribution; the intestinal subtype was observed in
280 Tamura et al
Surgery February 2015
Table I. Clinical characteristics of 13 patients with multiple MD intraductal papillary mucinous neoplasms of the pancreas Age at Patient Operation operation Sex 1
Duct type
MPD size Operative (mm) procedure
i ii i ii i ii i
66 72 76 79 60 62 75
M MD MD M MD MD M MD MD M a) MD, b) BD
ii
80 77 72 75
MD M MD 14 M MD 8 a) MD, b) BD
7
73
8
71
M a),b) MD, c) BD F a),b) MD
9 10
69 65
2 3 4
5 6
11
i ii
i
70
ii 12
78
13
72
16 14 18 14
10 6
Histologic grade
Histologic subtype
DP Inv-CA Rem TP HGD DP a),b) HGD Rem TP Inv-CA DP HGD Rem TP HGD DP a) Inv-CA, b) HGD Rem TP Inv-CA PD a),b) HGD DP HGD DP a) IGD, b) Inv-CA TP a),b),c) LGD
Gastric PB a),b) Intestinal Intestinal Intestinal Intestinal a) PB, b) Intestinal PB a),b) PB Intestinal a),b) Intestinal
TP
a) HGD, b) Inv-CA M a),b) MD 15 TP a),b) HGD M a),b) MD 30 TP a) Inv-CA, b) HGD a) HGD, M a) MD, b) BD 10 a) DP, b) PD b) LGD PDAC Rem DP PDAC F a),b) MD, 6.7 TP a),b) LGD, c) PDAC c) PDAC F a) MD, 9 TP a) Inv-CA, b) PDAC b) PDAC
Surveillance period (mo) Outcome 74
Alive
45
Alive
109
Death
104
Alive
126 48
Death Alive
a),b),c) Gastric
25
Alive
a) Gastric, b) PB a),b) Intestinal a),b) Intestinal
18
Death
41 211
Alive Alive
77
Alive
16
Alive
29
Death
a) Intestinal, b) Gastric — a), b) Gastric, c) — a) Gastric, b)—
BD, Branch duct; DP, distal pancreatectomy; F, female; HGD, high-grade dysplasia; Inv-CA, intraductal papillary mucinous neoplasm with an associated invasive carcinoma; IGD, intermediate-grade dysplasia; LGD, low-grade dysplasia; M, male; MD, main duct; Mono, monoclonal; MPD, main pancreatic duct; PB, pancreatobiliary; PDAC, pancreatic ductal adenocarcinoma; Poly, polyclonal; Rem, remnant; TP, total pancreatectomy; Wt, wild type.
47% of IPMNs harboring a GNAS mutation (17/36), but in 6% of the wild-type cohort (1/17). Conversely, the PB type was observed in only 6% of the GNAS mutant cohort (3/36) and in 35% of the wild-type cohort (6/17). According to a pairwise comparison of different histologic subtypes, GNAS mutations were more frequently observed in the intestinal subtype (94%; 17/18) than in other subtypes such as the gastric (67%; 14/21; P = .03) and PB subtypes (33%; 3/9; P < .01; Fig 3, A). However, there was no difference in the distribution of the histologic grade of dysplasia (P = .42; data not shown). The GNAS mutation patterns among the different histologic grades of IPMNs were also examined; 64% (7/11), 83% (5/6), 79% (15/19), and 56% (9/16) were GNAS mutants in the low-grade dysplasia, IGD, HGD, and Inv-CA cohorts, respectively (Fig 3, B). There was no difference in the cumulative diseasespecific survival rate (P = .92) between the GNAS wild-type and GNAS mutant cohorts (Fig 3, C).
DISCUSSION The main findings of the present study are as follows: (1) Distinct multiple segments of MDIPMNs might have monoclonal progression, (2) multiple MD-IPMNs were more frequently malignant (HGD/Inv-CA) than solitary MD-IPMNs, and (3) GNAS mutations were correlated with intestinal differentiation of MD-IPMNs, but this did not affect patient postoperative outcome. The clonality of MD-IPMNs has not been fully explored. We previously clarified the recurrence pattern after resection of MD-IPMNs using molecular analyses and assumed that recurrent lesions in the remnant pancreas after resection of MD-IPMNs were residual or skip areas of the initial tumor because the recurrent lesions had molecular characteristics similar to those of the initial MD-IPMNs.6 To determine whether synchronous and/or metachronous multisegmental MD-IPMNs are
Surgery Volume 157, Number 2
Fig 1. Flow diagram of the 70 MD-IPMN patients who underwent pancreatectomy in this study. In total, 70 patients were divided into 2 cohorts according to the pathologic mapping of the entire specimen after resection: Multiple (n = 13) and solitary (n = 57) cohorts. Finally, 13 patients in this series had multiple synchronous (n = 9) and/or metachronous (n = 6) neoplastic lesions. *Three patients had both synchronous and metachronous lesions in multiple areas. BD, Branch duct; MD, main duct; PDAC, pancreatic ductal adenocarcinoma.
actually monoclonal tumors, we carried out the present investigation in 13 patients with multiple lesions of MD-IPMN. The mutational patterns of multisegmental MD-IPMNs were consistent with each other in 10 of 11 patients examined, indicating that MD-IPMNs may actually be characterized by skip and monoclonal development in the MPD. Moreover, in patient 6, a BD-IPMN with Inv-CA developed in the remnant pancreas after partial pancreatectomy for multisegmental MDIPMNs with HGD and IGD. All 3 IPMNs stained positively for MUC2 and had the same mutational status, possibly suggesting that floating carcinoma cells are implanted not only into the MPD, but also into the BD. There were 5 patients who developed metachronous MD-IPMN in the remnant pancreas after partial pancreatectomy. Interestingly, all patients had undergone distal pancreatectomy for MD-IPMNs in the pancreatic body or tail at first operation, and recurrent lesions occurred in their remnant pancreatic head. Moreover, all the metachronous areas in these 5 patients had the same mutational findings,
Tamura et al 281
which supports our hypothesis that the floating neoplastic cells were implanted into the MPD of the pancreatic head along the natural flow of the pancreatic juice. However, simply examining KRAS/ GNAS mutations and histologic subtypes seems insufficient to prove genetic identity, and further investigation, such as a microarray analysis, is necessary to clarify this issue. The frequency of malignant IPMN (HGD and Inv-CA) was significantly higher in multisegmental MD-IPMNs, indicating that malignant MD-IPMNs might have a greater tendency to skip into the MPD than nonmalignant MD-IPMNs. Even if a negative pancreatic margin (noncarcinoma) can be obtained during partial pancreatectomy for MD-IPMN, it might be better to perform intraoperative irrigation cytology of the remnant pancreatic MPD as we and others did to rule out residual lesions in the remnant pancreas.11,12 As described previously, we consider that total pancreatectomy can be avoidable and partial pancreatectomy is acceptable if a negative resection margin (noncarcinoma) can be obtained.6 To protect against the occurrence of skip areas might be difficult; however, we believe that strict postoperative surveillance will lead to the early detection of metachronous skip lesions in the remnant pancreas. Yachida et al13 reported 2 types of somatic mutation in pancreatic cancer: Founder and progressor mutations. Founder mutations occur before the development of metastatic lesions. All other mutations are considered to be progressor mutations, which are associated with cancer metastasis. In our study, GNAS mutation was already found in 64% among low-grade IPMNs as others described,8,9 and the KRAS/GNAS mutational status was identical between the primary and metastatic sites in all 3 examined patients. These findings indicate that KRAS and GNAS mutations are founder mutations in the development of IPMNs. Mutant TP53, PIK3CA, and DPC4 are also reported as responsible genes although the frequency of these mutations is rare (Poly Poly Poly Mono Mono Mono Poly Mono Poly Poly
BD, Branch duct; Inv-CA, intraductal papillary mucinous neoplasm with an associated invasive carcinoma; MD, main duct; Mono, monoclonal; PB, pancreatobiliary; PDAC, pancreatic ductal adenocarcinoma; Poly, polyclonal; Wt, wild type.
permanent activation of the Gsa subunit, and mutant GNAS has been reported in several neoplasms.16 According to whole IPMN exome sequencing by the Johns Hopkins Institution group, as expected, both KRAS (81%) and GNAS (66%) mutations were frequently observed and all these mutation were located at codons 12 (KRAS) and 201 (GNAS), respectively.8,17 As a result, we focused only on exon 2 for KRAS and exon 8 for GNAS mutation in this study. Comprehensive analysis of the GNAS mutational status revealed no correlation between GNAS status and IPMN histologic grade or clinical characteristics, including patient postoperative outcomes. On the other hand, GNAS mutations are correlated with the intestinal subtype because 94% of intestinal type IPMNs had a GNAS mutation, whereas
only 33% of PB-type IPMNs had a GNAS mutation. These data are consistent with a recent study from the Johns Hopkins group,18 and GNAS mutational analysis might support the clarification of the mechanisms underlying initiation of different tumors. In addition, GNAS mutation analysis might be useful to distinguish concomitant PDAC from Inv-CA derived from IPMN, because concomitant PDAC rarely has GNAS mutations. Kanda et al19 demonstrated that duodenal collections obtained after secretin stimulation in patients with IPMNs had a mutant GNAS prevalence similar to that of the patients’ IPMNs. We previously reported that distinct PDACs frequently develop in the pancreas with gastric type IPMN without GNAS mutation.20 Hara et al21 showed that preoperative pancreatic juice cytology with MUC (mucin)
Surgery Volume 157, Number 2
Tamura et al 283
Fig 2. Schema of a patient with 2 distinct main duct intraductal papillary mucinous neoplasms (MD-IPMNs; patient 5). The mutational pattern and histologic subtype of these 2 distinct MD-IPMNs are consistent: Pancreatobiliary type IPMN with high-grade dysplasia and KRAS (G12D) mutation. The main pancreatic duct between the 2 distinct segments reveals normal epithelium without KRAS/GNAS mutation (upper middle panel; stain: hematoxylin and eosin). Wt, Wild type.
Fig 3. Comprehensive GNAS mutational analysis in 53 main duct intraductal papillary mucinous neoplasms. (A) Prevalence of GNAS codon 201 mutations according to the histologic subtype. GNAS mutation is observed in 67% of gastric, 94% of intestinal, and 33% of pancreatobiliary types. (B) Prevalence of GNAS codon 201 mutations according to the histologic grade. GNAS mutations are found in 64% of low-grade dysplasia (LGD), 83% of intermediate-grade dysplasia (IGD), 79% of high-grade dysplasia (HGD), and 56% of invasive carcinoma (Inv-CA). (C) There is no difference in the cumulative disease-specific survival rate between the GNAS WT and mutant cohorts. *Log-rank test.
staining was highly reliable for determination of the histologic subtype of IPMN. Additionally, evaluation of IPMN subtypes is considered to support
the prediction of prognosis.22-24 These reports indicate that pancreatic juice and/or duodenal fluid might provide useful information in the assessment
284 Tamura et al
of GNAS status and/or MUC expression. Thus, we might be able to predict the histologic subtype preoperatively and identify patients at high risk of developing concomitant PDAC. In conclusion, MD-IPMNs may be characterized by monoclonal skip progression. Even when a negative resection margin has been obtained, close attention should be paid to the potential for the metachronous occurrence of skip areas in the remnant pancreas after partial pancreatectomy for MD-IPMN. The authors thank Dr Junji Kishimoto of the Center for Clinical and Translational Research at Kyushu University Hospital, Fukuoka, Japan, for his special assistance in the statistical analyses of this study.
SUPPLEMENTARY DATA Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.surg.2014. 09.013. REFERENCES 1. Hruban RH, Pitman MB, Klimstra DS. Tumors of the pancreas. In: Silverberg SG, Sobin LH, editors. Atlas of tumor pathology, AFIP fourth series, fascicle 6. Washington, DC: American Registry of Pathology; 2007. p. 75-110. 2. Tanaka M, Fernandez-del Castillo C, Adsay V, Chari S, Falconi M, Jang JY, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012;12:183-97. 3. Izawa T, Obara T, Tanno S, Mizukami Y, Yanagawa N, Kohgo Y. Clonality and field cancerization in intraductal papillarymucinous tumors of the pancreas. Cancer 2001;92:1807-17. 4. Z’graggen K, Rivera JA, Compton CC, Pins M, Werner J, Fernandez-del Castillo C, et al. Prevalence of activating Kras mutations in the evolutionary stages of neoplasia in intraductal papillary mucinous tumors of the pancreas. Ann Surg 1997;226:491-8. 5. Matthaei H, Norris AL, Tsiatis AC, Olino K, Hong SM, dal Molin M, et al. Clinicopathological characteristics and molecular analyses of multifocal intraductal papillary mucinous neoplasms of the pancreas. Ann Surg 2012;255: 326-33. 6. Tamura K, Ohtsuka T, Ideno N, Aso T, Shindo K, Aishima S, et al. Treatment strategy for main duct intraductal papillary mucinous neoplasms of the pancreas based on the assessment of recurrence in the remnant pancreas after resection: a retrospective review. Ann Surg 2014; 259:360-8. 7. Fl ejou JF. Barrett’s oesophagus: from metaplasia to dysplasia and cancer. Gut 2004;54:i6-12. 8. Wu J, Matthaei H, Maitra A, Dal Molin M, Wood LD, Eshleman JR, et al. Recurrent GNAS mutations define an unexpected pathway for pancreatic cyst development. Sci Transl Med 2011;3:92ra66. 9. Furukawa T, Kuboki Y, Tanji E, Yoshida S, Hatori T, Yamamoto M, et al. Whole-exome sequencing uncovers frequent GNAS mutations in intraductal papillary mucinous neoplasms of the pancreas. Sci Rep 2011;1:161.
Surgery February 2015
10. Adsay NV, Kloppel G, Fukushima N, Offerhaus GJA, Furukawa T, Pitman MB, et al. Intraductal papillary-mucinous neoplasms of the pancreas. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, editors. World Health Organization classification of tumors, pathology and genetics of tumors of the digestive system. Lyon, France: IARC Press; 2010. p. 304-13. 11. Eguchi H, Ishikawa O, Ohigashi H, Sasaki Y, Yamada T, Nakaizumi A, et al. Role of intraoperative cytology combined with histology in detecting continuous and skip type intraductal cancer existence for intraductal papillary mucinous carcinoma of the pancreas. Cancer 2006;107:2567-75. 12. Mori Y, Ohtsuka T, Tamura K, Ideno N, Aso T, Kono H, et al. Intraoperative irrigation cytology of the remnant pancreas to detect remnant distinct pancreatic ductal adenocarcinoma in patients with intraductal papillary mucinous neoplasm undergoing partial pancreatectomy. Surgery 2014;155:67-73. 13. Yachida S, Jones S, Bozic I, Antal T, Leary R, Fu B, et al. Distant metastasis occurs late during the genetic evolution of pancreatic cancer. Nature 2010;467:1114-7. 14. Lubezky N, Ben-Haim M, Marmor S, Brazowsky E, Rechavi G, Klausner JM, et al. High-throughput mutation profiling in intraductal papillary mucinous neoplasm (IPMN). J Gastrointest Surg 2011;15:503-11. 15. Shi C, Hruban RH. Intraductal papillary mucinous neoplasm. Hum Pathol 2012;43:1-16. 16. Nault JC, Fabre M, Couchy G, Pilati C, Jeannot E, Tran Van Nhieu J, et al. GNAS-activating mutations define a rare subgroup of inflammatory liver tumors characterized by STAT3 activation. J Hepatol 2012;56:184-91. 17. Wu J, Jiao Y, Dal Molin M, Maitra A, de Wilde RF, Wood LD, et al. Whole-exome sequencing of neoplastic cysts of the pancreas reveals recurrent mutations in components of ubiquitin-dependent pathways. Proc Natl Acad Sci U S A 2011;108:21188-93. 18. Dal Molin M, Matthaei H, Wu J, Blackford A, Debeljak M, Rezaee N, et al. Clinicopathological correlates of activating GNAS mutations in intraductal papillary mucinous neoplasm (IPMN) of the pancreas. Ann Surg Oncol 2013;20:3802-8. 19. Kanda M, Knight S, Topazian M, Syngal S, Farrell J, Lee J, et al. Mutant GNAS detected in duodenal collections of secretin-stimulated pancreatic juice indicates the presence or emergence of pancreatic cysts. Gut 2013;62:1024-33. 20. Ideno N, Ohtsuka T, Kono H, Fujiwara K, Oda Y, Aishima S, et al. Intraductal papillary mucinous neoplasms of the pancreas with distinct pancreatic ductal adenocarcinomas are frequently of gastric subtype. Ann Surg 2013;258:141-51. 21. Hara T, Ikebe D, Odaka A, Sudo K, Nakamura K, Yamamoto H, et al. Preoperative histological subtype classification of intraductal papillary mucinous neoplasms (IPMN) by pancreatic juice cytology with MUC stain. Ann Surg 2013; 257:1103-11. 22. Mino-Kenudson M, Fernandez-del Castillo C, Baba Y, Valsangkar NP, Liss AS, Hsu M, et al. Prognosis of invasive intraductal papillary mucinous neoplasm depends on histological and precursor epithelial subtypes. Gut 2011; 60:1712-20. 23. Sadakari Y, Ohuchida K, Nakata K, Ohtsuka T, Aishima S, Takahata S, et al. Invasive carcinoma derived from the nonintestinal type intraductal papillary mucinous neoplasm of the pancreas has a poorer prognosis than that derived from the intestinal type. Surgery 2010;147:812-7. 24. Distler M, Kersting S, Niedergethmann M, Aust DE, Franz M, R€ uckert F, et al. Pathohistological subtype predicts survival in patients with intraductal papillary mucinous neoplasm (IPMN) of the pancreas. Ann Surg 2013;258:324-30.
Background. Main duct intraductal papillary mucinous neoplasms (MD-IPMNs) may occur in 1 or multiple segments of the pancreatic duct. Unlike multifocal branch duct (BD)-IPMNs, the clonality of multisegmental MD-IPMNs remains unclear. GNAS mutations are common and specific for IPMNs, and mutational assessment might be useful to determine the clonality of IPMNs as well as to detect high-risk IPMN with distinct ductal adenocarcinoma (pancreatic ductal adenocarcinoma [PDAC]). Our aim was to clarify clonality using GNAS status in multisegmental MD-IPMNs. Methods. We retrospectively reviewed the medical records of 70 patients with MD-IPMN. Histologic subtypes and KRAS/GNAS mutations were investigated, and the clonal relationships among multisegmental MD-IPMNs were assessed. Mutational analysis was performed using high-resolution melting analysis and subsequent Sanger/pyrosequencing. Results. Thirteen patients had multiple synchronous and/or metachronous lesions. Seven of these 13 patients had multiple MD-IPMNs; 3 had multiple MD-IPMNs and distinct BD-IPMNs; 1 had multiple MD-IPMNs and a distinct PDAC; 1 had a solitary MD-IPMN, BD-IPMN, and PDAC; and 1 had a solitary MD-IPMN and PDAC. KRAS/GNAS mutations were consistent in 10 of 11 multisegmental MD-IPMNs, whereas MD-IPMNs, BD-IPMNs, and PDACs tended to show different mutational patterns. The frequency of malignant IPMNs was significantly higher in the multisegment cohort; malignant IPMNs constituted 90% (9/10) of the multiple cohort and 56% (32/57) of the solitary cohort (P = .04). Mutant GNAS was more frequently observed in the intestinal subtype (94%) than the others. Conclusion. MD-IPMNs can be characterized by monoclonal skip progression. Close attention should be paid to the possible presence of skip areas during or after partial pancreatectomy. (Surgery 2015;157:277-84.) From the Departments of Surgery and Oncology,a Anatomic Pathology,b Medicine and Bioregulatory Science,c and Clinical Radiology,d Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan
Conflicts of interest and source of funding: The authors declare no conflicts of interest. This study was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Numbers 24390318, 24390319, 23390327, 25293285, 25670586, 25670585, 25670584, 25670582, and 24592030. Accepted for publication September 10, 2014. Reprint requests: Masao Tanaka, MD, PhD, FACS, Department of Surgery and Oncology, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 8128582, Japan. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2014.09.013
SEVERAL STUDIES have reported that intraductal papillary mucinous neoplasms of the pancreas (IPMNs) are characterized by multifocality, the prevalence of which varies from 20 to 41%.1,2 Most previous studies assessed both main duct (MD-IPMNs) and branch duct IPMNs (BD-IPMNs), and the majority of multifocal IPMNs were of the BD type. A field carcinogenesis effect seems to play an important role in some cases of IPMN.3 Z’graggen et al4 reported that KRAS mutations were often detected in the histologically normal epithelium of some IPMN patients, which seems to demonstrate field malignant transformation in the entire pancreas SURGERY 277
278 Tamura et al
of IPMN patients. On the basis of such genetic alterations, several studies examining multifocal lesions of IPMN in the same pancreas showed different patterns of KRAS mutations or loss of heterozygosity, suggesting that multifocal IPMNs may originate from independent precursor lesions.3,5 Conversely, we previously suggested that MDIPMNs might demonstrate monoclonal progression because the mutational patterns of recurrent lesions in the remnant pancreas were consistent with those of the initial lesions in 4 MD-IPMN patients, although the surgical margin was negative for neoplastic cells.6 This finding seems to differ from the polyclonality of BD-IPMNs. MD-IPMN may occur in 1 or multiple segments of the main pancreatic duct; however, the clonality of the genetic alterations of multisegmental MD-IPMNs remains unclear. Several epithelial malignancies reportedly arise through an adenoma–carcinoma sequence.7 IPMNs are well known to show a wide spectrum of histologic differentiation from adenomas and borderline neoplasms to carcinoma, and are thus considered to be a good model of the adenoma– carcinoma sequence. GNAS is a gene encoding the guanine nucleotide-binding protein a subunit. Recent reports have stated that GNAS mutations are common and specific for IPMNs, whereas other pancreatic neoplasms, including pancreatic ductal adenocarcinoma (PDAC), rarely have GNAS mutations.8,9 Wu et al8 found that 47.8% of low-grade IPMNs had GNAS mutations, indicating that GNAS mutation might be associated with tumor initiation as a founder mutation of IPMNs. In this study, the mutation findings and histologic subtypes among distinct synchronous/metachronous multiple MD segments and the GNAS mutational status among MD-IPMNs were comprehensively analyzed to clarify the clonal evolution and genetic alterations among MD-IPMNs. PATIENTS AND METHODS The present study was approved by the Ethics Committee of Kyushu University and conducted according to the Ethical Guidelines for Human Genome/Gene Research enacted by the Japanese Government and the Declaration of Helsinki. We reviewed retrospectively the medical records of 70 patients who had been histologically diagnosed with MD-IPMN after resection at the Department of Surgery and Oncology, Kyushu University Hospital between 1987 and 2013. The data of the initial 56 patients were described in our previous study for different purposes.6 MD-IPMN was
Surgery February 2015 defined as >5-mm segmental or diffuse dilation of the main pancreatic duct (MPD) without >5mm dilation of the BD.2 Mixed-type IPMN was included as an MD-IPMN in this study, as previously reported.6 In accordance with the 2010 World Health Organization criteria, IPMN components were classified into 4 different histologic grades: Lowgrade dysplasia, intermediate-grade dysplasia (IGD), high-grade dysplasia (HGD), and invasive carcinoma (Inv-CA).10 IPMNs were also classified into 4 histologic subtypes: Gastric, intestinal, pancreatobiliary (PB), and oncocytic. When multiple IPMNs were simultaneously present in the same pancreas, the histologic grade and subtype were determined according to of the IPMN with the highest degree of dysplasia. All patients were classified into 1 of 2 cohorts according to the pathologic mapping of the whole specimen after resection: Solitary and multiple cohorts. A solitary MD-IPMN was defined as 1 continuous tumor, and a multisegmental MDIPMN was defined as a tumor affecting $2 distinct segments of the MPD separated from each other by an uninvolved segment of normal MPD. We confirmed that the MPD epithelium between the 2 distinct lesions had no papillary dysplastic changes. Additionally, there were 2 types of multiple lesions: Synchronous and metachronous tumors. Synchronous lesions were defined as multiple lesions that were resected during the same operation, and metachronous lesions were defined as distinct lesions that had metachronously developed in the remnant pancreas after partial pancreatectomy with negative surgical margins. KRAS and GNAS mutations and histologic subtypes of each lesion were investigated to assess the clonality of multisegmental MD-IPMNs. For the purpose of mutational analysis, appropriate tissue blocks were obtained from formalinfixed, paraffin-embedded tissue samples. Tumor tissues were dissected macroscopically if the neoplastic cellularity was $70%. They were microdissected with a sterile needle under an optical microscope (BX41 TF; Olympus, Tokyo, Japan) in cases of low neoplastic cellularity or for the normal pancreatic duct epithelium. Genomic DNA was subsequently extracted using the QIAamp DNA Formalin-Fixed, Paraffin-Embedded Tissue Kit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). The mutational status of exon 2 of KRAS and exon 8 of GNAS was investigated with high-resolution melting curve analysis followed by Sanger sequencing and/or pyrosequencing. Briefly, polymerase chain
Surgery Volume 157, Number 2
reactions for high-resolution melting curve analysis were carried out in a final volume of 10 mL, composed of 5 mL of SsoFast EvaGreen Supermix (Bio-Rad Laboratories, Hercules, CA); 3.2 mL of diethylpyrocarbonate-treated, pyrogen-free, DNase/RNase-free water (Invitrogen, Carlsbad, CA); 1 mL (10 ng/mL) of template DNA; and 0.4 mL (10 pmol/mL) of each forward and reverse primer pair. All samples were tested in triplicate. Polymerase chain reaction products that were indicated to have mutations by highresolution melting curve analysis were subjected to Sanger sequencing to confirm the single-nucleotide polymorphism hot spot as previously described.6 If a mutation was not detected by Sanger sequencing, additional pyrosequencing with the PyroMark Q24 (Qiagen) using PyroMark Gold reagents (Qiagen) was performed according to the manufacturer’s instructions. When single nucleotide polymorphisms were not detected by pyrosequencing, the allele was defined as wild type. The primers used in this study are listed in the Supplementary Table. Statistical analyses were completed by JMP statistical software (version 9.0.2; SAS Institute, Cary, NC). The Chi-square test or Fisher’s exact test was used to evaluate differences in categorical data between 2 cohorts, and the Mann–Whitney U test was used for continuous data. The disease-specific survival rate was analyzed using the Kaplan–Meier method with log-rank univariate comparisons. RESULTS Clinicopathologic characteristics. In total, 70 resected MD-IPMNs including 13 mixed IPMNs were reviewed. The histologic grades of these 70 IPMNs at the time of the initial operation were lowgrade dysplasia in 20 patients (29%), IGD in 7 (10%), HGD in 19 (27%), Inv-CA in 23 (33%; including 7 T1a, formerly called minimally invasive, carcinomas), and unknown in 1 (HGD or Inv-CA). The histologic subtypes were gastric type in 32 patients (47%), intestinal in 24 (35%), PB in 9 (13%), oncocytic in 3 (4%), and unknown in 2. Three PDACs (2 synchronous and 1 metachronous) were simultaneously present with MD-IPMNs, as previously reported.5 Seven patients developed metachronous MD-IPMNs in the remnant pancreas after partial pancreatectomy, 4 of which were free from neoplastic cells at the pancreatic margin and 3 of which were positive for atypical cells or had unknown details. There was no difference in either the cumulative disease-specific survival rate (P = .44; Supplementary Figure) or the rate of recurrence in the remnant pancreas (P = .96; Supplementary Figure) between patients with (n = 22) and without (n = 36) atypical cells at the resected margin.
Tamura et al 279
Analyses of KRAS and GNAS mutations and histologic subtype of multiple lesions in MD-IPMNs. Thirteen patients in this series had multiple synchronous (n = 9) and/or metachronous (n = 6) neoplastic lesions (Table I). Seven of these 13 patients had multiple-segment MD-IPMNs (patients 1–3, 5, and 8–10); 3 had multiple MD-IPMNs and a distinct BD-IPMN; 1 had multiple MD-IPMNs and a distinct PDAC; 1 had a solitary MD-IPMN, BD-IPMN, and PDAC; and 1 had a solitary MD-IPMN and distinct PDAC (Fig 1). In 10 of the 11 patients with multiple MD-IPMNs (excluding patient 7), the mutational patterns of the multiple segments of the MD-IPMNs were consistent with one another in the same pancreas (Table II; Fig 2). In patients 1 and 8, the histologic subtype of the main lesion differed from that of a sublesion (one was gastric and the other was PB type), although they seemed to be histologically similar. MD-IPMN and BD-IPMN lesions tended to show different mutational patterns and/or histologic subtypes, indicating a polyclonal relationship, as observed in patients 4, 7, and 11. There were also different mutational patterns between MD-IPMNs and concomitant PDACs. Additionally, when the histologic grade distribution was compared between multiple (n = 10) and solitary (n = 57) IPMN cohorts (excluding 3 patients with concomitant PDACs), the frequency of malignant IPMNs (HGD/Inv-CA) was significantly higher in the cohort with multiple segment MD-IPMNs; malignant IPMNs constituted 90% (9/10) of the multiple cohort and 56% (32/57) of the solitary cohort (P = .04). Effects of GNAS mutation on clinicopathologic features. A comprehensive GNAS mutational analysis of 53 MD-IPMNs resected between 2000 and 2013 was performed to investigate the significance of GNAS mutation in the biological behavior of MD-IPMN. Seventeen samples were excluded for such reasons as old formalin-fixed, paraffinembedded samples (obtained before 1999), inadequate DNA quality, insufficient amplification of the polymerase chain reaction product, and lack of adequate tissue samples. Thirty-six of 53 MD-IPMNs (68%) harbored a GNAS codon 201 point mutation, and 17 (32%) were determined to have wild-type GNAS. Clinical characteristics, such as patient age, symptoms, MPD diameter, presence of mural nodules, serum levels of carcinoembryonic antigen and carbohydrate antigen 19-9 (normal or high), and histologic grade and subtype, were compared between the GNAS wild-type and mutant cohorts. There was a significant difference in the subtype distribution; the intestinal subtype was observed in
280 Tamura et al
Surgery February 2015
Table I. Clinical characteristics of 13 patients with multiple MD intraductal papillary mucinous neoplasms of the pancreas Age at Patient Operation operation Sex 1
Duct type
MPD size Operative (mm) procedure
i ii i ii i ii i
66 72 76 79 60 62 75
M MD MD M MD MD M MD MD M a) MD, b) BD
ii
80 77 72 75
MD M MD 14 M MD 8 a) MD, b) BD
7
73
8
71
M a),b) MD, c) BD F a),b) MD
9 10
69 65
2 3 4
5 6
11
i ii
i
70
ii 12
78
13
72
16 14 18 14
10 6
Histologic grade
Histologic subtype
DP Inv-CA Rem TP HGD DP a),b) HGD Rem TP Inv-CA DP HGD Rem TP HGD DP a) Inv-CA, b) HGD Rem TP Inv-CA PD a),b) HGD DP HGD DP a) IGD, b) Inv-CA TP a),b),c) LGD
Gastric PB a),b) Intestinal Intestinal Intestinal Intestinal a) PB, b) Intestinal PB a),b) PB Intestinal a),b) Intestinal
TP
a) HGD, b) Inv-CA M a),b) MD 15 TP a),b) HGD M a),b) MD 30 TP a) Inv-CA, b) HGD a) HGD, M a) MD, b) BD 10 a) DP, b) PD b) LGD PDAC Rem DP PDAC F a),b) MD, 6.7 TP a),b) LGD, c) PDAC c) PDAC F a) MD, 9 TP a) Inv-CA, b) PDAC b) PDAC
Surveillance period (mo) Outcome 74
Alive
45
Alive
109
Death
104
Alive
126 48
Death Alive
a),b),c) Gastric
25
Alive
a) Gastric, b) PB a),b) Intestinal a),b) Intestinal
18
Death
41 211
Alive Alive
77
Alive
16
Alive
29
Death
a) Intestinal, b) Gastric — a), b) Gastric, c) — a) Gastric, b)—
BD, Branch duct; DP, distal pancreatectomy; F, female; HGD, high-grade dysplasia; Inv-CA, intraductal papillary mucinous neoplasm with an associated invasive carcinoma; IGD, intermediate-grade dysplasia; LGD, low-grade dysplasia; M, male; MD, main duct; Mono, monoclonal; MPD, main pancreatic duct; PB, pancreatobiliary; PDAC, pancreatic ductal adenocarcinoma; Poly, polyclonal; Rem, remnant; TP, total pancreatectomy; Wt, wild type.
47% of IPMNs harboring a GNAS mutation (17/36), but in 6% of the wild-type cohort (1/17). Conversely, the PB type was observed in only 6% of the GNAS mutant cohort (3/36) and in 35% of the wild-type cohort (6/17). According to a pairwise comparison of different histologic subtypes, GNAS mutations were more frequently observed in the intestinal subtype (94%; 17/18) than in other subtypes such as the gastric (67%; 14/21; P = .03) and PB subtypes (33%; 3/9; P < .01; Fig 3, A). However, there was no difference in the distribution of the histologic grade of dysplasia (P = .42; data not shown). The GNAS mutation patterns among the different histologic grades of IPMNs were also examined; 64% (7/11), 83% (5/6), 79% (15/19), and 56% (9/16) were GNAS mutants in the low-grade dysplasia, IGD, HGD, and Inv-CA cohorts, respectively (Fig 3, B). There was no difference in the cumulative diseasespecific survival rate (P = .92) between the GNAS wild-type and GNAS mutant cohorts (Fig 3, C).
DISCUSSION The main findings of the present study are as follows: (1) Distinct multiple segments of MDIPMNs might have monoclonal progression, (2) multiple MD-IPMNs were more frequently malignant (HGD/Inv-CA) than solitary MD-IPMNs, and (3) GNAS mutations were correlated with intestinal differentiation of MD-IPMNs, but this did not affect patient postoperative outcome. The clonality of MD-IPMNs has not been fully explored. We previously clarified the recurrence pattern after resection of MD-IPMNs using molecular analyses and assumed that recurrent lesions in the remnant pancreas after resection of MD-IPMNs were residual or skip areas of the initial tumor because the recurrent lesions had molecular characteristics similar to those of the initial MD-IPMNs.6 To determine whether synchronous and/or metachronous multisegmental MD-IPMNs are
Surgery Volume 157, Number 2
Fig 1. Flow diagram of the 70 MD-IPMN patients who underwent pancreatectomy in this study. In total, 70 patients were divided into 2 cohorts according to the pathologic mapping of the entire specimen after resection: Multiple (n = 13) and solitary (n = 57) cohorts. Finally, 13 patients in this series had multiple synchronous (n = 9) and/or metachronous (n = 6) neoplastic lesions. *Three patients had both synchronous and metachronous lesions in multiple areas. BD, Branch duct; MD, main duct; PDAC, pancreatic ductal adenocarcinoma.
actually monoclonal tumors, we carried out the present investigation in 13 patients with multiple lesions of MD-IPMN. The mutational patterns of multisegmental MD-IPMNs were consistent with each other in 10 of 11 patients examined, indicating that MD-IPMNs may actually be characterized by skip and monoclonal development in the MPD. Moreover, in patient 6, a BD-IPMN with Inv-CA developed in the remnant pancreas after partial pancreatectomy for multisegmental MDIPMNs with HGD and IGD. All 3 IPMNs stained positively for MUC2 and had the same mutational status, possibly suggesting that floating carcinoma cells are implanted not only into the MPD, but also into the BD. There were 5 patients who developed metachronous MD-IPMN in the remnant pancreas after partial pancreatectomy. Interestingly, all patients had undergone distal pancreatectomy for MD-IPMNs in the pancreatic body or tail at first operation, and recurrent lesions occurred in their remnant pancreatic head. Moreover, all the metachronous areas in these 5 patients had the same mutational findings,
Tamura et al 281
which supports our hypothesis that the floating neoplastic cells were implanted into the MPD of the pancreatic head along the natural flow of the pancreatic juice. However, simply examining KRAS/ GNAS mutations and histologic subtypes seems insufficient to prove genetic identity, and further investigation, such as a microarray analysis, is necessary to clarify this issue. The frequency of malignant IPMN (HGD and Inv-CA) was significantly higher in multisegmental MD-IPMNs, indicating that malignant MD-IPMNs might have a greater tendency to skip into the MPD than nonmalignant MD-IPMNs. Even if a negative pancreatic margin (noncarcinoma) can be obtained during partial pancreatectomy for MD-IPMN, it might be better to perform intraoperative irrigation cytology of the remnant pancreatic MPD as we and others did to rule out residual lesions in the remnant pancreas.11,12 As described previously, we consider that total pancreatectomy can be avoidable and partial pancreatectomy is acceptable if a negative resection margin (noncarcinoma) can be obtained.6 To protect against the occurrence of skip areas might be difficult; however, we believe that strict postoperative surveillance will lead to the early detection of metachronous skip lesions in the remnant pancreas. Yachida et al13 reported 2 types of somatic mutation in pancreatic cancer: Founder and progressor mutations. Founder mutations occur before the development of metastatic lesions. All other mutations are considered to be progressor mutations, which are associated with cancer metastasis. In our study, GNAS mutation was already found in 64% among low-grade IPMNs as others described,8,9 and the KRAS/GNAS mutational status was identical between the primary and metastatic sites in all 3 examined patients. These findings indicate that KRAS and GNAS mutations are founder mutations in the development of IPMNs. Mutant TP53, PIK3CA, and DPC4 are also reported as responsible genes although the frequency of these mutations is rare (Poly Poly Poly Mono Mono Mono Poly Mono Poly Poly
BD, Branch duct; Inv-CA, intraductal papillary mucinous neoplasm with an associated invasive carcinoma; MD, main duct; Mono, monoclonal; PB, pancreatobiliary; PDAC, pancreatic ductal adenocarcinoma; Poly, polyclonal; Wt, wild type.
permanent activation of the Gsa subunit, and mutant GNAS has been reported in several neoplasms.16 According to whole IPMN exome sequencing by the Johns Hopkins Institution group, as expected, both KRAS (81%) and GNAS (66%) mutations were frequently observed and all these mutation were located at codons 12 (KRAS) and 201 (GNAS), respectively.8,17 As a result, we focused only on exon 2 for KRAS and exon 8 for GNAS mutation in this study. Comprehensive analysis of the GNAS mutational status revealed no correlation between GNAS status and IPMN histologic grade or clinical characteristics, including patient postoperative outcomes. On the other hand, GNAS mutations are correlated with the intestinal subtype because 94% of intestinal type IPMNs had a GNAS mutation, whereas
only 33% of PB-type IPMNs had a GNAS mutation. These data are consistent with a recent study from the Johns Hopkins group,18 and GNAS mutational analysis might support the clarification of the mechanisms underlying initiation of different tumors. In addition, GNAS mutation analysis might be useful to distinguish concomitant PDAC from Inv-CA derived from IPMN, because concomitant PDAC rarely has GNAS mutations. Kanda et al19 demonstrated that duodenal collections obtained after secretin stimulation in patients with IPMNs had a mutant GNAS prevalence similar to that of the patients’ IPMNs. We previously reported that distinct PDACs frequently develop in the pancreas with gastric type IPMN without GNAS mutation.20 Hara et al21 showed that preoperative pancreatic juice cytology with MUC (mucin)
Surgery Volume 157, Number 2
Tamura et al 283
Fig 2. Schema of a patient with 2 distinct main duct intraductal papillary mucinous neoplasms (MD-IPMNs; patient 5). The mutational pattern and histologic subtype of these 2 distinct MD-IPMNs are consistent: Pancreatobiliary type IPMN with high-grade dysplasia and KRAS (G12D) mutation. The main pancreatic duct between the 2 distinct segments reveals normal epithelium without KRAS/GNAS mutation (upper middle panel; stain: hematoxylin and eosin). Wt, Wild type.
Fig 3. Comprehensive GNAS mutational analysis in 53 main duct intraductal papillary mucinous neoplasms. (A) Prevalence of GNAS codon 201 mutations according to the histologic subtype. GNAS mutation is observed in 67% of gastric, 94% of intestinal, and 33% of pancreatobiliary types. (B) Prevalence of GNAS codon 201 mutations according to the histologic grade. GNAS mutations are found in 64% of low-grade dysplasia (LGD), 83% of intermediate-grade dysplasia (IGD), 79% of high-grade dysplasia (HGD), and 56% of invasive carcinoma (Inv-CA). (C) There is no difference in the cumulative disease-specific survival rate between the GNAS WT and mutant cohorts. *Log-rank test.
staining was highly reliable for determination of the histologic subtype of IPMN. Additionally, evaluation of IPMN subtypes is considered to support
the prediction of prognosis.22-24 These reports indicate that pancreatic juice and/or duodenal fluid might provide useful information in the assessment
284 Tamura et al
of GNAS status and/or MUC expression. Thus, we might be able to predict the histologic subtype preoperatively and identify patients at high risk of developing concomitant PDAC. In conclusion, MD-IPMNs may be characterized by monoclonal skip progression. Even when a negative resection margin has been obtained, close attention should be paid to the potential for the metachronous occurrence of skip areas in the remnant pancreas after partial pancreatectomy for MD-IPMN. The authors thank Dr Junji Kishimoto of the Center for Clinical and Translational Research at Kyushu University Hospital, Fukuoka, Japan, for his special assistance in the statistical analyses of this study.
SUPPLEMENTARY DATA Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.surg.2014. 09.013. REFERENCES 1. Hruban RH, Pitman MB, Klimstra DS. Tumors of the pancreas. In: Silverberg SG, Sobin LH, editors. Atlas of tumor pathology, AFIP fourth series, fascicle 6. Washington, DC: American Registry of Pathology; 2007. p. 75-110. 2. Tanaka M, Fernandez-del Castillo C, Adsay V, Chari S, Falconi M, Jang JY, et al. International consensus guidelines 2012 for the management of IPMN and MCN of the pancreas. Pancreatology 2012;12:183-97. 3. Izawa T, Obara T, Tanno S, Mizukami Y, Yanagawa N, Kohgo Y. Clonality and field cancerization in intraductal papillarymucinous tumors of the pancreas. Cancer 2001;92:1807-17. 4. Z’graggen K, Rivera JA, Compton CC, Pins M, Werner J, Fernandez-del Castillo C, et al. Prevalence of activating Kras mutations in the evolutionary stages of neoplasia in intraductal papillary mucinous tumors of the pancreas. Ann Surg 1997;226:491-8. 5. Matthaei H, Norris AL, Tsiatis AC, Olino K, Hong SM, dal Molin M, et al. Clinicopathological characteristics and molecular analyses of multifocal intraductal papillary mucinous neoplasms of the pancreas. Ann Surg 2012;255: 326-33. 6. Tamura K, Ohtsuka T, Ideno N, Aso T, Shindo K, Aishima S, et al. Treatment strategy for main duct intraductal papillary mucinous neoplasms of the pancreas based on the assessment of recurrence in the remnant pancreas after resection: a retrospective review. Ann Surg 2014; 259:360-8. 7. Fl ejou JF. Barrett’s oesophagus: from metaplasia to dysplasia and cancer. Gut 2004;54:i6-12. 8. Wu J, Matthaei H, Maitra A, Dal Molin M, Wood LD, Eshleman JR, et al. Recurrent GNAS mutations define an unexpected pathway for pancreatic cyst development. Sci Transl Med 2011;3:92ra66. 9. Furukawa T, Kuboki Y, Tanji E, Yoshida S, Hatori T, Yamamoto M, et al. Whole-exome sequencing uncovers frequent GNAS mutations in intraductal papillary mucinous neoplasms of the pancreas. Sci Rep 2011;1:161.
Surgery February 2015
10. Adsay NV, Kloppel G, Fukushima N, Offerhaus GJA, Furukawa T, Pitman MB, et al. Intraductal papillary-mucinous neoplasms of the pancreas. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, editors. World Health Organization classification of tumors, pathology and genetics of tumors of the digestive system. Lyon, France: IARC Press; 2010. p. 304-13. 11. Eguchi H, Ishikawa O, Ohigashi H, Sasaki Y, Yamada T, Nakaizumi A, et al. Role of intraoperative cytology combined with histology in detecting continuous and skip type intraductal cancer existence for intraductal papillary mucinous carcinoma of the pancreas. Cancer 2006;107:2567-75. 12. Mori Y, Ohtsuka T, Tamura K, Ideno N, Aso T, Kono H, et al. Intraoperative irrigation cytology of the remnant pancreas to detect remnant distinct pancreatic ductal adenocarcinoma in patients with intraductal papillary mucinous neoplasm undergoing partial pancreatectomy. Surgery 2014;155:67-73. 13. Yachida S, Jones S, Bozic I, Antal T, Leary R, Fu B, et al. Distant metastasis occurs late during the genetic evolution of pancreatic cancer. Nature 2010;467:1114-7. 14. Lubezky N, Ben-Haim M, Marmor S, Brazowsky E, Rechavi G, Klausner JM, et al. High-throughput mutation profiling in intraductal papillary mucinous neoplasm (IPMN). J Gastrointest Surg 2011;15:503-11. 15. Shi C, Hruban RH. Intraductal papillary mucinous neoplasm. Hum Pathol 2012;43:1-16. 16. Nault JC, Fabre M, Couchy G, Pilati C, Jeannot E, Tran Van Nhieu J, et al. GNAS-activating mutations define a rare subgroup of inflammatory liver tumors characterized by STAT3 activation. J Hepatol 2012;56:184-91. 17. Wu J, Jiao Y, Dal Molin M, Maitra A, de Wilde RF, Wood LD, et al. Whole-exome sequencing of neoplastic cysts of the pancreas reveals recurrent mutations in components of ubiquitin-dependent pathways. Proc Natl Acad Sci U S A 2011;108:21188-93. 18. Dal Molin M, Matthaei H, Wu J, Blackford A, Debeljak M, Rezaee N, et al. Clinicopathological correlates of activating GNAS mutations in intraductal papillary mucinous neoplasm (IPMN) of the pancreas. Ann Surg Oncol 2013;20:3802-8. 19. Kanda M, Knight S, Topazian M, Syngal S, Farrell J, Lee J, et al. Mutant GNAS detected in duodenal collections of secretin-stimulated pancreatic juice indicates the presence or emergence of pancreatic cysts. Gut 2013;62:1024-33. 20. Ideno N, Ohtsuka T, Kono H, Fujiwara K, Oda Y, Aishima S, et al. Intraductal papillary mucinous neoplasms of the pancreas with distinct pancreatic ductal adenocarcinomas are frequently of gastric subtype. Ann Surg 2013;258:141-51. 21. Hara T, Ikebe D, Odaka A, Sudo K, Nakamura K, Yamamoto H, et al. Preoperative histological subtype classification of intraductal papillary mucinous neoplasms (IPMN) by pancreatic juice cytology with MUC stain. Ann Surg 2013; 257:1103-11. 22. Mino-Kenudson M, Fernandez-del Castillo C, Baba Y, Valsangkar NP, Liss AS, Hsu M, et al. Prognosis of invasive intraductal papillary mucinous neoplasm depends on histological and precursor epithelial subtypes. Gut 2011; 60:1712-20. 23. Sadakari Y, Ohuchida K, Nakata K, Ohtsuka T, Aishima S, Takahata S, et al. Invasive carcinoma derived from the nonintestinal type intraductal papillary mucinous neoplasm of the pancreas has a poorer prognosis than that derived from the intestinal type. Surgery 2010;147:812-7. 24. Distler M, Kersting S, Niedergethmann M, Aust DE, Franz M, R€ uckert F, et al. Pathohistological subtype predicts survival in patients with intraductal papillary mucinous neoplasm (IPMN) of the pancreas. Ann Surg 2013;258:324-30.
