Pancreatic cancer is a common malignant tumor of

Transcription

1 PDVR Original in treatment Article / of Pancreas pancreatic cancer Pancreaticoduodenectomy with portal vein/superior mesenteric vein resection for patients with pancreatic cancer with venous invasion Wei-Lin Wang, Song Ye, Sheng Yan, Yan Shen, Min Zhang, Jian Wu and Shu-Sen Zheng Hangzhou, China BACKGROUND: With the development of new surgical techniques, pancreaticoduodenectomy (PD) with portal vein or superior mesenteric vein (PV/SMV) resection has been used in the treatment of patients with borderline resectable pancreatic cancer. However, opinions of surgeons differ in the effectiveness of this surgical technique. This study aimed to investigate the effectiveness of this approach in patients with pancreatic cancer. METHODS: Follow-up visits and retrospective analysis were carried out of 208 patients with pancreatic cancer who had undergone PD (PD group) and PD combined with PV/SMV resection and reconstruction (PDVR group) from June 2009 to May 2013 at our center. Statistical analysis was performed to compare the clinical features, the difference of survival time and risk factors of venous invasion in pancreatic cancer. Factors relating to postoperative survival time of pancreatic cancer were also investigated. RESULTS: In the PDVR group, which consisted of 42 cases, the 1-, 2- and 3-year survival rates were 70%, 41% and 16%, respectively and the median survival time was 20.0 months. Among the 166 patients in the PD group, the 1-, 2- and 3-year survival rates were 80%, 52%, and 12%, respectively with the median survival time of 26.0 months. No significant difference in survival time and R0 resection ratio was found between the two groups. Lumbodorsal pain, tumor with pancreatic capsular invasion and bile duct infiltration were found to be independent risk factors for PV invasion in pancreatic cancer. In addition, non R0 resection, Author Affiliations: Division of Hepatobiliary Pancreatic Surgery, First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou , China (Wang WL, Ye S, Yan S, Shen Y, Zhang M, Wu J and Zheng SS) Corresponding Author: Shu-Sen Zheng, MD, PhD, FACS, Division of Hepatobiliary Pancreatic Surgery, First Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou , China (Tel/Fax: ; shusenzheng@zju.edu.cn) 2015, Hepatobiliary Pancreat Dis Int. All rights reserved. doi: /S (15) Published online July 15, large tumor size (>2 cm) and poorly differentiated tumor were independent risk factors for survival time in post-pd. CONCLUSIONS: The tumor has a higher chance of venous invasion if preoperative imagings indicate that it juxtaposes with the vessel. Lumbodorsal pain is the chief complaint. Patients with pancreatic cancer associated with PV involvement should receive PDVR for R0 resection when preoperational assessment shows the chance for eradication. (Hepatobiliary Pancreat Dis Int 2015;14: ) KEY WORDS: pancreatic cancer; pancreaticoduodenectomy; portal vein resection; superior mesenteric vein Introduction Pancreatic cancer is a common malignant tumor of the digestive system, with pancreatic ductal adenocarcinoma as the most common subtype. It ranks 13th and 8th in incidence and mortality respectively in malignant tumors. [1] Pancreatic cancer is known to have the poorest prognosis among abdominal tumors with a 5-year survival rate less than 5% [2] and surgical resection by far provides the best outcome. Operative mortality is less than 5% and the 5-year survival rate goes up to 15%-25%. [3-7] However, suitable candidates for surgical resection at time of diagnosis account for only 10%-20% of all patients. [8] Nearly 30% of the patients are diagnosed at borderline resectable stages when the tumors have already invaded the portal vein or superior mesenteric vein (PV/SMV). For years, resection has been considered a contraindication for tumors invading the PV/ SMV. Moore et al [9] first reported the application of pancreaticoduodenectomy (PD) with partial resection of the SMV in the treatment of pancreatic cancer in Later, PD in combination with PV resection was performed in tumors of the pancreatic head, the inferior common bile duct and ampulla vater infiltrating the PV/SMV. This ap- Hepatobiliary Pancreat Dis Int,Vol 14,No 4 August 15,

2 Hepatobiliary & Pancreatic Diseases International proach is not widely accepted due to the high incidence of surgical complications, mortality, and uncertainty of therapeutic efficacy. After Fortner et al [10] reported the use of regional pancreatectomy, Japan and some Western countries implemented similar extended resection measures, with a combination of vein resection and reconstruction as the main method. However, the use of PD with vein resection in pancreatic cancer patients with PV/SMV invasion still remains controversial. Some studies [11-14] showed that PD with vein resection and reconstruction is safe and feasible; the tumor resection rate can be improved without increasing the incidence of complication and mortality; others, however, did not validate this notion. Han et al [15] reported that even though radical surgery is the only way to treat patients, its effectiveness may be very low when pancreatic cancer has invaded the PV/SMV, especially when the tunica intima is involved. In this study, we wish to verify the prognostic implications (either in terms of postoperative complications or long-term survival) in a contemporary series of patients who underwent PD with and without vein resection. of type C, 23 of type D, and 11 of type E. Postoperative pathology of patients in the PDVR group showed PV/ SMV invasion. Preoperative assessment and surgical procedure of curative resection The surgical approach for treating pancreatic cancer was evaluated by the following steps (Fig. 1). The method of vein resection and reconstruction was determined based on the results of preoperative imaging evaluation and actual condition of intraoperation. The methods adopted include: (i) direct end-to-end anastomosis after vein resection, which was applied in 28 patients with resection of veins less than 4 cm, with no tension occurring after anastomosis; (ii) bilateral end-toend anastomosis, which was applied to patients with vein resection longer than 4 cm and allograft blood vessel had to be used as substitute after resection (Fig. 2A); (iii) separate end-to-end anastomosis, which was used when Methods General data of patients A retrospective analysis was made in 208 patients diagnosed as pancreatic head cancer between June 2009 and May 2013 at the First Affiliated Hospital of Zhejiang University School of Medicine in Hangzhou, China. Among them, 166 underwent PD (PD group) and 42 underwent PD combined with PV/SMV resection and reconstruction (PDVR group). Inclusion criteria: (i) patients did not have metastatic or locally advanced tumor at the time of diagnosis; (ii) patients had resectable and borderline resectable tumors based on preoperative imaging analysis according to the latest National Comprehensive Cancer Network (NCCN) guideline; (iii) postoperative pathology confirmed pancreatic ductal adenocarcinoma; (iv) PDVR was based on preoperative imaging and intraoperative findings of PV/SMV involvement with venous narrowing and/or vein abutment by the tumor. Of the 166 patients in the PD group, 115 were males and 51 females. They were years old, with an average age of 60.5 years. Of the 42 patients in the PDVR group, 26 were males and 16 females. They were years old, with an average age of 59.4 years. Preoperative evaluation was performed on all patients using thin-slice spiral contrast-enhanced CT. Based on the Loyer classification, [16] venous infiltration of pancreatic cancer was divided into six types (A-F). Patients were either type A or B in the PD group, while the PDVR group consisted of 8 patients Fig. 1. Schematic drawing of preoperative evaluation of method for tumor resection. PD: pancreaticoduodenectomy; PD+VR: pancreaticoduodenectomy plus vein resection and reconstruction; SMV: superior mesenteric vein; PV: portal vein; SMA: superior mesenteric artery; CTA: computed tomographyic angiography; MRA: magnetic resonance angiography. Fig. 2. Schematic drawing of PV/SMV resection and reconstruction methods. 430 Hepatobiliary Pancreat Dis Int,Vol 14,No 4 August 15,2015

3 PDVR in treatment of pancreatic cancer the convergence of the SMV, splenic and portal vein was invaded by the tumor. The vein was then substituted by a Y-shaped allograft blood vessel (Fig. 2B and C); (iv) substitution by Y-shaped allograft blood vessels with endto-side anastomosis, which was employed when tumor invasion occurred from the convergence downwards to the bifurcation of the SMV (Fig. 2D). Portal veins or iliac arteries from liver transplant donors were used as blood vessel allografts. Blood type matching was ensured in all patients. Fresh harvested blood vessels less than 24 hours were used whenever possible. Frozen blood vessels were used when fresh blood vessels were unavailable. Enbloc resection was used in both PD and PDVR groups in accordance with the tumor-free principle. Whipple's approach was combined with Child's reconstruction. Lymph node dissection was performed conventionally at stations 13 and 17 (already resected along with the specimens), lymph nodes at stations 8, 9 and 12, as well as station 14 group on the right side of the superior mesenteric artery. Finally, pancreaticojejunal anastomosis, bilioenteric anastomosis and gastrointestinal anastomosis were made in sequence. One or two drainage tubes were inserted posterior to the sites of bilioenteric anastomosis and pancreaticojejunal anastomosis. Postoperative treatment and follow-up Patients did not receive conventional anticoagulant therapy after vein resection and reconstruction in our study. Venous blood flow was evaluated one week after surgery using Doppler B-ultrasound. The condition of blood vessels was evaluated one month after surgery by abdominal computed tomographic angiography (CTA). The definition of postoperative complications was adopted from the International Study Group of Pancreatic Surgery (ISGPS). [17] Special personnel were hired for a 3-month regular telephone follow-up. All subsequent treatments, survival time, relapse and quality of life of the patients were investigated. The date of last follow-up was December, Statistical analysis Statistical analysis was carried out using SPSS 19.0 (IBM, Armonk, NY, USA). Clinical features and complications were compared using the Chi-square test and Student's t test. Independent risk factors for venous invasion were analyzed by logistic regression. Survival rate was determined by the life table method, and the survival curve was analyzed by the Kaplan-Meier method. A correlation analysis was made between the survival time and each of the following factors in 208 patients by the Kaplan-Meier method: gender, age (>60 years), smoking, drinking, lumbodorsal pain, body weight loss (>5 kg), diabetes, postoperative parameters included pancreatic leakage, intra-abdominal bleeding, gastroparesis, abdominal infection, degree of tumor differentiation, R0 resection, tumor size (>2 cm), pancreatic capsular invasion, bile duct infiltration, lymph node metastasis, and TNM staging. Independent risk factors associated with survival time were analyzed using Cox regression. A P<0.05 was considered statistically significant. Results Clinicopathologic features and complications The general clinical features and incidence of complications of the two groups as well as the inter-group comparison are shown in Table 1. All patients in the PDVR group were pathologically confirmed as having PV invasion, lumbodorsal pain, pancreatic capsular invasion, bile duct infiltration and degree of tumor differentiation. These were considered putative risk factors for PV invasion. Multiple logistic regressions were performed to analyze these risk factors. Results indicated that lumbodorsal pain, pancreatic capsular invasion and bile duct infiltration were independent risk factors for PV invasion in pancreatic cancer (Table 2). Moreover, the incidences of major postoperative complications were similar between the two groups (P>0.05, Table 1). Survival rate analysis The 30- and 90-day mortality rates of the 42 patients in the PDVR group were 2.4% (1/42) and 2.4% (1/42), respectively. Among the 166 patients in the PD group, the 30- and 90-day mortality rates were 1.2% (2/166) and 3.0% (5/166), respectively (P>0.05, Table 1). The overall 1-, 2- and 3-year survival rates were 78%, 50% and 12%, respectively, with a median of 24.0 months (95% CI: , Fig. 3). The 1-, 2- and 3-year survival rates of the PDVR group were 70%, 41% and 16%, respectively, with a median of 20.0 months (95% CI: ). In the PD group, the 1-, 2- and 3-year survival rates were 80%, 52% and 12%, respectively, with a median of 26.0 months (95% CI: ). A comparison of survival curves between the two groups is shown in Fig. 4. No significant difference in survival time was found (P=0.205). Analysis of factors influencing survival time after PD and PDVR Poorly differentiated tumor, non R0 resection, tumor size (>2 cm), pancreatic capsular invasion, lymph node metastasis, and advanced TNM stage were all adverse factors for survival time in pancreatic cancer patients who underwent resection (Fig. 5). Cox regression analy- Hepatobiliary Pancreat Dis Int,Vol 14,No 4 August 15,

4 Hepatobiliary & Pancreatic Diseases International Table 1. Comparison of general clinical features and complications between the PDVR and PD groups Variables PDVR group (n=42) PD group (n=166) P value Age (yr) 59.4± ± Gender Male Female History of smoking Yes No History of drinking Yes No Lumbodorsal pain Yes No Body weight loss (kg) > Diabetes Yes 5 34 No CA19-9 (IU/mL) 941.0± ± Operation time (min) 352± ± Intraoperative blood loss (ml) 520± ± Tumor size (cm) > Pancreatic capsular invasion Yes No 2 79 Bile duct infiltration Yes No Tumor differentiation Poorly Moderately 9 60 Well 1 14 R0 resection Yes No 8 36 Lymph node metastasis Yes No Postoperative pancreatic fistula Yes 4 10 No Postoperative delayed gastric emptying Yes 8 24 No Postoperative hemorrhage Yes 3 7 No Postoperative abdominal infection Yes 1 9 No day mortality Dead 1 2 Alive day mortality Dead 1 5 Alive Table 2. Multiple factor logistic regression analysis of the risk factors of venous invasion Risk factors Odds ratio P value 95% CI Lumbodorsal pain Pancreatic capsular invasion Bile duct infiltration Table 3. Cox regression analysis of multiple factors affecting the survival time in pancreatic cancer Risk factors Relative risk P value 95% CI Tumor size (>2 cm) Poorly differentiation R0 resection Fig. 3. Analysis of overall survival of 208 patients by the Kaplan- Meier method. Fig. 4. Comparison of survival curves between the PDVR and PD groups using the Kaplan-Meier method (P=0.205). sis showed that non R0 resection, large tumor size (>2 cm) and poorly differentiated tumors were independent risk factors for survival time after resection (Table 3). The median survival time of patients treated by R0 resection was 26.0 months and the other was 7.0 months, correspondingly (P=0.000). 432 Hepatobiliary Pancreat Dis Int,Vol 14,No 4 August 15,2015

5 PDVR in treatment of pancreatic cancer Fig. 5. Single factor analysis of the prognosis of pancreatic cancer by the Kaplan-Meier method showing poorly differentiated tumor (A), non R0 resection (B), tumor size (>2 cm) (C), pancreatic capsular invasion (D), lymph node metastasis (E) and advanced TNM stage (F), are risk factors for survival time. Discussion The PDVR approach for radical resection of pancreatic cancers has increasingly been adopted at medical centers with the advance of the surgical technique. Radical resection of the tumor is considered the best curative option for patients with middle-stage and advanced pancreatic cancer. Generally speaking, conditions suitable for the PDVR approach are: (i) no distant metastasis or peritoneal seeding; (ii) no superior mesenteric artery, hepatic artery or celiac artery invasion; (iii) invasion or blockage of PV/SMV by the tumor and a possibility for safe vein reconstruction. It is necessary to preoperatively evaluate the vascular invasion. This assessment is mainly based on imaging data, especially angiography (CTA or MRA). The invasive digital subtraction angiography technique is no longer the first choice. In clinical practice, it is difficult to distinguish tumor infiltration from inflammatory adhesion based on imaging results alone. Therefore, we tried to find the other approaches to detect vascular invasion. The present study found that lumbodorsal pain, Hepatobiliary Pancreat Dis Int,Vol 14,No 4 August 15,

6 Hepatobiliary & Pancreatic Diseases International pancreatic capsular invasion and bile duct infiltration are independent risk factors for PV invasion. However, pancreatic capsular invasion and bile duct infiltration are not easy to identify before surgery. Based on our study, we proposed that blood vessel invasion should be highly suspected when the patient has lumbodorsal pain and the images indicate that the tumor juxtaposes with the vessel. R0 resection is the object of PDVR and its ratio is about 40%-70% in all PDVR cases. [18, 19] Data from the literature showed that the 5-year survival rate of the patients after R0 resection is about 15%-25%. [3-6] The difference is rooted from the technical level of the medical centers and the criteria of R0 resection. In our study, the R0 resection ratio of the PDVR group and PD group was 81.0% (34/42) and 78.3% (130/166), respectively (P=0.708). Our data demonstrated that the R0 resection ratio of the PDVR group was similar to that of the PD group. Kulemann's study showed that the the R0 resection ratio of the PDVR group was better than that of the PD group. [20] These data indicated that the radical resection of pancreatic cancer with PV invasion is similar to the routine procedure of the pancreatic cancer resection in patients without vascular invasion. And this can be the precondition and basis of PDVR in pancreatic cancer treatment. For patients in whom preoperative evaluation indicates vessels invasion with certainty, en-bloc resection with blood vessels is preferred to achieve R0 resection according to the tumor-free principle. However, this approach remains controversial because of perioperative risk and long-time prognosis. In the current study, allograft blood vessels from liver transplant donors were used to replace the invaded vessels. This technique has seldom been reported in previous studies. No statistically significant difference was found in survival time between the two groups. Likewise, no significant difference was found between the two groups in terms of intraoperative blood loss and operation time. The most common perioperative complications were pancreatic leakage, abdominal bleeding, gastroparesis and abdominal infection. There were no significant differences between the two groups in terms of complications. Just consistent with us, Adham et al [21] recently shared their experience in treating pancreatic cancer by PD with blood vessel resection at two hepatopancreatobiliary centers in Europe. Their data showed that the 1-, 2-, and 3-year disease-free survival rates of patients who underwent blood vessel resection were 36.0%, 15.0% and 12.0%, respectively. The median disease-free survival time was 8.7 months. The 1-, 2- and 3-year overall survival rates were 56.6%, 28.9% and 19.2%, respectively, with an median of 14.2 months. Thus, compared with PD, PDVR did not show significant increase in difficulty of the procedure or incidence of perioperative complications. Furthermore, the medium and long-term survival rate and median survival rate did not differ significantly between the two methods. Matsuno and coworkers found that the typical survival time is 3-6 months in patients with pancreatic head tumor with blood vessels invasion and without surgical intervention. [22] This indicated that for patients suitable for blood vessel resection and reconstruction, PDVR can be effective in prolonging the survival time. Therefore, in the experienced center, PDVR is a safe and effective treatment strategy for patients with PV/SMV involvement. The proper length of the portal vein to be resected for direct anastomosis remains a topic of much debate. It is generally believed that a resection length of 3-5 cm is sufficient to prevent tension in the vessel after direct anastomosis. However, some surgeons suggested that the permissible range for direct anastomosis is 5-8 cm by dissociation between the mesentery and falciform ligament of the liver. Based on our experiences in organ transplantation and the general condition of allograft blood vessels from the donor, the appropriate resection length is generally 4 cm. When the length of blood vessel resected was greater than 4 cm, it was substituted by allograft blood vessels. Compared with artificial blood vessels used previously at our center, the use of allograft vessels has led to a significant reduction in the rate of early thrombus formation. Hence, our results indicated that allograft blood vessel transplant is a feasible option for patients undergoing PDVR. Furthermore, survival analysis was performed for each factor affecting survival time in pancreatic cancer. It was found that the poorly differentiated tumor, non R0 resection, tumor size (>2 cm), pancreatic capsular invasion, lymph node metastasis and advanced TNM stage were adverse factors for postoperative survival time. Among them, non R0 resection, tumor size (>2 cm) and poorly differentiated tumor were independent risk factors. In the three independent risk factors of pancreatic cancer, tumor size and tumor differentiation were the features of the tumor, and R0 resection was the only factor manageable and therefore, R0 resection was the only pursuable factor in the treatment of pancreatic cancer. In conclusion, apart from imaging data, lumbodorsal pain is helpful in identifying pancreatic cancer patients with portal vein invasion. These patients can be treated with PDVR to get the same R0 resection ratio and similar outcomes as those without vessel invasion. Furthermore, the long-term survival after PD or PDVR surgery was independently influenced by R0 resection, tumor size and tumor differentiation. We concluded that the patients 434 Hepatobiliary Pancreat Dis Int,Vol 14,No 4 August 15,2015

7 PDVR in treatment of pancreatic cancer diagnosed with pancreatic cancer accompanied by PV involvement should receive PDVR for R0 resection when preoperational assessment allows us to do so. Contributors: ZSS proposed the study. WWL, YeS, YS, SY, ZM and WJ performed research and wrote the first draft. WWL and YeS collected and analyzed the data, and contributed equally to the article. All authors contributed to the design and interpretation of the study and to further drafts. ZSS is the guarantor. Funding: This study was supported by grants from the Research Special Fund for Public Welfare Industry of Health ( ), the National Key Specialty Construction of Clinical Project (General Surgery), the Education Department of Zhejiang Province (Y ) and the Health Department of Zhejiang Province ( ). Ethical approval: This study was approved by the Institutional Ethics Committee of First Affiliated Hospital of Zhejiang University School of Medicine. Written informed consent was first obtained from the patient or a family member. Competing interest: No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. References 1 Ferlay J, Shin HR, Bray F, Forman D, Mathers C, Parkin DM. Estimates of worldwide burden of cancer in 2008: GLOBO- CAN Int J Cancer 2010;127: Jemal A, Siegel R, Ward E, Hao Y, Xu J, Murray T, et al. Cancer statistics, CA Cancer J Clin 2008;58: Loos M, Kleeff J, Friess H, Büchler MW. Surgical treatment of pancreatic cancer. Ann N Y Acad Sci 2008;1138: Cameron JL, Riall TS, Coleman J, Belcher KA. One thousand consecutive pancreaticoduodenectomies. Ann Surg 2006;244: Richter A, Niedergethmann M, Sturm JW, Lorenz D, Post S, Trede M. Long-term results of partial pancreaticoduodenectomy for ductal adenocarcinoma of the pancreatic head: 25-year experience. World J Surg 2003;27: Carpelan-Holmström M, Nordling S, Pukkala E, Sankila R, Lüttges J, Klöppel G, et al. Does anyone survive pancreatic ductal adenocarcinoma? A nationwide study re-evaluating the data of the Finnish Cancer Registry. Gut 2005;54: Winter JM, Cameron JL, Campbell KA, Arnold MA, Chang DC, Coleman J, et al pancreaticoduodenectomies for pancreatic cancer: A single-institution experience. J Gastrointest Surg 2006;10: Sener SF, Fremgen A, Menck HR, Winchester DP. Pancreatic cancer: a report of treatment and survival trends for patients diagnosed from , using the National Cancer Database. J Am Coll Surg 1999;189: Moore GE, Sako Y, Thomas LB. Radical pancreatoduodenectomy with resection and reanastomosis of the superior mesenteric vein. Surgery 1951;30: Fortner JG, Kim DK, Cubilla A, Turnbull A, Pahnke LD, Shils ME. Regional pancreatectomy: en bloc pancreatic, portal vein and lymph node resection. Ann Surg 1977;186: Ramacciato G, Mercantini P, Petrucciani N, Giaccaglia V, Nigri G, Ravaioli M, et al. Does portal-superior mesenteric vein invasion still indicate irresectability for pancreatic carcinoma? Ann Surg Oncol 2009;16: Nakao A, Kanzaki A, Fujii T, Kodera Y, Yamada S, Sugimoto H, et al. Correlation between radiographic classification and pathological grade of portal vein wall invasion in pancreatic head cancer. Ann Surg 2012;255: Raut CP, Tseng JF, Sun CC, Wang H, Wolff RA, Crane CH, et al. Impact of resection status on pattern of failure and survival after pancreaticoduodenectomy for pancreatic adenocarcinoma. Ann Surg 2007;246: Tseng JF, Raut CP, Lee JE, Pisters PW, Vauthey JN, Abdalla EK, et al. Pancreaticoduodenectomy with vascular resection: margin status and survival duration. J Gastrointest Surg 2004;8: Han SS, Park SJ, Kim SH, Cho SY, Kim YK, Kim TH, et al. Clinical significance of portal-superior mesenteric vein resection in pancreatoduodenectomy for pancreatic head cancer. Pancreas 2012;41: Loyer EM, David CL, Dubrow RA, Evans DB, Charnsangavej C. Vascular involvement in pancreatic adenocarcinoma: reassessment by thin-section CT. Abdom Imaging 1996;21: Tan WJ, Kow AW, Liau KH. Moving towards the New International Study Group for Pancreatic Surgery (ISGPS) definitions in pancreaticoduodenectomy: a comparison between the old and new. HPB (Oxford) 2011;13: Chua TC, Saxena A. Extended pancreaticoduodenectomy with vascular resection for pancreatic cancer: a systematic review. J Gastrointest Surg 2010;14: Siriwardana HP, Siriwardena AK. Systematic review of outcome of synchronous portal-superior mesenteric vein resection during pancreatectomy for cancer. Br J Surg 2006;93: Kulemann B, Hoeppner J, Wittel U, Glatz T, Keck T, Wellner UF, et al. Perioperative and long-term outcome after standard pancreaticoduodenectomy, additional portal vein and multivisceral resection for pancreatic head cancer. J Gastrointest Surg 2015;19: Adham M, Mirza DF, Chapuis F, Mayer AD, Bramhall SR, Coldham C, et al. Results of vascular resections during pancreatectomy from two European centres: an analysis of survival and disease-free survival explicative factors. HPB (Oxford) 2006;8: Matsuno S, Egawa S, Fukuyama S, Motoi F, Sunamura M, Isaji S, et al. Pancreatic Cancer Registry in Japan: 20 years of experience. Pancreas 2004;28: Received July 19, 2014 Accepted after revision April 2, 2015 Hepatobiliary Pancreat Dis Int,Vol 14,No 4 August 15,