[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address Please contact the publisher to request reinstatement.
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Views 1,181
Citations 0
Brief Report
Resident's Forum
June 2014

Retrograde vs Conventional Dissection Technique in PancreaticoduodenectomyA Pilot Study

Author Affiliations
  • 1Department of General Surgery, Changzheng Hospital, Second Military Medical University, Shanghai, People’s Republic of China
JAMA Surg. 2014;149(6):604-607. doi:10.1001/jamasurg.2013.3399

Importance  A retrograde dissection technique of pancreaticoduodenectomy in a caudocranial direction has been described recently.

Observations  Fifteen consecutive patients who underwent retrograde pancreaticoduodenectomy were compared with 15 consecutive patients operated on through a conventional approach. The mean (SD) intraoperative blood loss was 407 (202) mL in the retrograde group compared with 423 (253) mL in the conventional group (P = .84). The mean (SD) operative duration was 255 (57) minutes in the retrograde group compared with 264 (54) minutes in the conventional group (P = .66). The overall morbidity was 7 of 15 patients (47%) in the retrograde group and 6 of 15 (40%) in the conventional group (P > .99). Neither group had a positive resection margin or a perioperative death.

Conclusions and Relevance  The retrograde dissection technique had no significant difference in perioperative outcomes compared with the conventional dissection technique and could serve as an alternative dissection approach in pancreaticoduodenectomy.

Pancreaticoduodenectomy (PD) is increasingly being performed in experienced centers. With better anesthesia, technical refinements, and improved perioperative care, morbidity and mortality have declined remarkably over the past few decades.1 However, it remains a challenging operation.2 Due to the complex and technically demanding nature of this procedure, a number of modifications have emerged over the years. Most modifications focus on reconstruction, notably pancreatic anastomosis.3 Little attention has been paid to the technique of resection. Standard resection of PD is usually performed in a craniocaudal direction. Recently, Hackert et al4 described a retrograde dissection technique in a caudocranial direction. This technique differs from the conventional dissection technique, since the resection is performed in a retrograde way, starting with the uncinate process. Shrikhande et al5 reported that a combined artery-first and uncinate process–first approach was comparable with the uncinate process–first approach only regarding perioperative outcomes in PD (similar to that described by Hackert et al). However, they highlighted the specific value of adopting the artery-first approach and did not document the technique and advantage of the uncinate process–first approach in detail. As stated by Hackert et al, studies are needed to evaluate the possible advantages of the retrograde vs the conventional dissection technique regarding operative parameters and postoperative outcomes. With this background, we conducted a pilot study to compare the perioperative outcomes of the retrograde and conventional dissection techniques during the performance of a PD.


In this prospective comparative study from August 1, 2011, to July 30, 3012, a total of 15 consecutive patients who underwent PD with the retrograde dissection technique were compared with 15 consecutive patients who had PD with the conventional dissection technique. The indication for PD was the presence of a periampullary tumor. Consecutive patients were included to obviate any possible selection bias. All patients were operated on by the same surgeon (Z.-Q.H.), who specializes in pancreatic surgery. The study was reviewed and approved by the Committee on Ethics of Biomedicine Research, Second Military Medical University, Shanghai. Written informed consent was obtained from the patients involved in the study.

Retrograde Dissection Technique

Briefly, the retrograde dissection technique differs from the conventional dissection technique in resection order. After Kocher mobilization, the hepatoduodenal ligament is prepared, including lymphadenectomy and exposure of the gastroduodenal artery and common bile duct, which can be cut at this point after removal of the gallbladder. The next step is division and translocation of the first jejunal loop to the right side of the mesenteric root. Afterward, retrograde resection of the pancreatic head is performed, starting with the uncinate process in a caudocranial direction (eFigure 1 [Supplement]). This is followed by transection of the lower portion of the stomach. As a result, the specimen is attached only to the pancreas body itself (eFigure 2 [Supplement]). Transection of the pancreas neck is performed as the last step of resection.

Data Collection

Preoperative demographic data, clinical history, examination and investigation findings, details of the surgical procedures, and all relevant perioperative variables, including the postoperative course and complications, and final histopathologic diagnosis were recorded in a prospectively maintained database.

Statistical Analysis

All continuous data were expressed as mean (SD). Categorical variables were compared using the χ2 test (or Fisher exact test when appropriate). The t test or nonparametric tests were used for continuous data according to their distribution. P < .05 was considered statistically significant.


Thirty consecutive patients who underwent PD for periampullary tumors were included in this study. The first 15 patients were operated on using the conventional dissection technique (conventional group), while in the remaining 15 patients, a retrograde dissection technique was used as described earlier (retrograde group). Twenty (67%) patients were men and 10 (33%) were women. The mean age was 57.1 (9.9) years. The 2 groups were comparable regarding age, body mass index, sex, preoperative type 2 diabetes mellitus, preoperative biliary drainage, preoperative hemoglobin, and other organ resection (Table 1).

Table 1.  
Clinical Characteristics and Intraoperative Outcomesa
Clinical Characteristics and Intraoperative Outcomesa

The mean intraoperative blood loss was 407 (202) mL in the retrograde group compared with 423 (253) mL in the conventional group (P = .84). The mean volume of blood transfused was 133 (279) mL (3 patients; range, 0-800 mL) in the retrograde group and 93.3 (249) mL (2 patients; range, 0-800 mL) in the conventional group (P = .68). The mean operative duration was 255 (57) minutes in the retrograde group compared with 264 (54) minutes in the conventional group (P = .66) (Table 1).

There was no perioperative mortality in either group. The overall morbidity was 7 of 15 patients (47%) in the retrograde group and 6 of 15 (40%) in the conventional group (P > .99). Details of the postoperative course and complications of both groups are shown in Table 2. Both groups were comparable regarding postoperative complications. The mean postoperative stay was 16.8 (8.2) days in the retrograde group compared with 15.6 (7.8) days in the conventional group (P = .68). There was no readmission within 30 days after surgery in either group.

Table 2.  
Postoperative Course and Complicationsa
Postoperative Course and Complicationsa

Table 3 provides details of postoperative pathologic outcomes. Both groups were comparable regarding tumor type, size, differentiation, and TNM staging. The mean number of lymph nodes (LNs) retrieved was 10.0 (6.2) in the retrograde group compared with 9.5 (5.7) in the conventional group (P = .82). The mean number of metastatic LNs retrieved was 0.7 (1.1) in the retrograde group compared with 0.7 (2.0) in the conventional group (P = .99). Also, there was no positive resection margin in either group.

Table 3.  
Postoperative Pathologic Outcomesa
Postoperative Pathologic Outcomesa

From our experience, the retrograde dissection technique has 3 potential advantages. First, bleeding control is easy to achieve. The initial mobilization of the first jejunal loop and transposition to the right side allow wide mobility, with the advantage of having an excellent overview of all small vessels with the chance to selectively clip or ligate them. Moreover, manual control of the pancreatic head by the surgeon’s left hand can help to create tension, which facilitates the dissection and avoids any accidental injury. Second, it offers the opportunity to completely lateralize the uncinate process to the right side and dissect it from the retroperitoneum and superior mesenteric vessels under direct vision. Thus, the surgical margins are clearly viewed throughout the whole resection stage, which could possibly result in a favorable oncologic outcome. Last, such a technique is particularly useful when the tumors involve the main portal vein or, when portal vein resection is anticipated, the portomesenteric junction. With such a technique, portal vein anastomosis could be conducted after removing the specimen, providing sufficient room to conduct a meticulous and precise end-to-end anastomosis.

Several preoperative factors are associated with operative difficulty and could perhaps influence perioperative outcomes, such as body mass index, preoperative biliary drainage, and anemia.68 In this study, both groups were comparable regarding these factors. Therefore, preoperative characteristics would not confound the outcomes.

The mean operative duration in both groups of patients showed no significant difference. Since the overall operative duration of a PD is affected by a number of factors and steps, time savings by using the retrograde dissection technique probably are not reflected in our series of patients. One reason might be that in the present study, retrograde dissection is a new technique that is time-consuming until surgeons become familiar with it.

The mean blood loss and volume of blood transfused in both groups showed no significant difference. Furthermore, there was no significant difference in the overall and specific morbidity between the 2 groups. None of the complications required further surgical intervention, and all were successfully treated by conservative therapy. These results imply that the retrograde dissection technique is as safe as the conventional dissection technique.

As seen in Table 3, the retrograde dissection technique can be used in all indications for PD and is not limited to any special anatomical situations. In addition, the present study demonstrates that tumors up to 5 cm can be safely handled by the retrograde dissection technique.

Another concern regarding the use of the retrograde dissection technique is the adequacy of LN retrieval and the risk of positivity of the retroperitoneal margin, which are significantly associated with local recurrence and impair long-term survival.9 We did not perform an extended lymphadenectomy because of the lack of evidence for improved survival but instead removed LNs around the head of the pancreas, common hepatic artery, and hepatoduodenal ligament.10 The present study showed that the number of LNs retrieved was not different between the 2 groups. In addition, none of the patients in both groups had positive margins on final histopathologic examination.

Postoperative hospital stay was not different between the 2 groups. This is not surprising since a prolonged hospital stay is mostly caused by the complications of pancreatic fistula and delayed gastric emptying, which were not influenced by the dissection technique.11


Since perioperative outcomes are comparable to the conventional method of dissection, the retrograde dissection technique can now be considered a viable option in modern PD to supplement the standard technique. Future larger multicenter long-term studies may help to evaluate the oncologic outcomes of the retrograde vs conventional dissection technique.

Section Editor: Richard D. Schulick, MD, MBA; Pamela A. Lipsett, MD, MPHE.
Back to top
Article Information

Accepted for Publication: June 24, 2013.

Corresponding Author: Zhi-Qian Hu, MD, Department of General Surgery, Changzheng Hospital, Second Military Medical University, No. 415, Fengyang Road, Shanghai, 200003, People’s Republic of China (czhuzq@aliyun.com).

Published Online: April 23, 2014. doi:10.1001/jamasurg.2013.3399.

Author Contributions: Dr Zhou had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Zhou, Y. Wang, Hu.

Acquisition, analysis, or interpretation of data: Zhou, Zhang, Ruan, W.-J. Wang, Sun, Hu.

Drafting of the manuscript: Zhou.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Zhou, Zhang.

Administrative, technical, or material support: Y. Wang, Ruan, W.-J. Wang, Sun.

Study supervision: Hu.

Conflict of Interest Disclosures: None reported.

Correction: This article was corrected on June 16, 2014, to fix a typographical error in the title and Abstract.

Schmidt  CM, Turrini  O, Parikh  P,  et al.  Effect of hospital volume, surgeon experience, and surgeon volume on patient outcomes after pancreaticoduodenectomy: a single-institution experience. Arch Surg. 2010;145(7):634-640.PubMedArticle
Are  C, Dhir  M, Ravipati  L.  History of pancreaticoduodenectomy: early misconceptions, initial milestones and the pioneers. HPB (Oxford). 2011;13(6):377-384.PubMedArticle
Yang  SH, Dou  KF, Sharma  N, Song  WJ.  The methods of reconstruction of pancreatic digestive continuity after pancreaticoduodenectomy: a meta-analysis of randomized controlled trials. World J Surg. 2011;35(10):2290-2297.PubMedArticle
Hackert  T, Werner  J, Weitz  J, Schmidt  J, Büchler  MW.  Uncinate process first—a novel approach for pancreatic head resection. Langenbecks Arch Surg. 2010;395(8):1161-1164.PubMedArticle
Shrikhande  SV, Barreto  SG, Bodhankar  YD,  et al.  Superior mesenteric artery first combined with uncinate process approach versus uncinate process first approach in pancreatoduodenectomy: a comparative study evaluating perioperative outcomes. Langenbecks Arch Surg. 2011;396(8):1205-1212.PubMedArticle
Shimoda  M, Katoh  M, Yukihiro  I, Kita  J, Sawada  T, Kubota  K.  Body mass index is a risk factor of pancreatic fistula after pancreaticoduodenectomy. Am Surg. 2012;78(2):190-194.PubMed
Coates  JM, Beal  SH, Russo  JE,  et al.  Negligible effect of selective preoperative biliary drainage on perioperative resuscitation, morbidity, and mortality in patients undergoing pancreaticoduodenectomy. Arch Surg. 2009;144(9):841-847.PubMedArticle
Hughes  C, Hurtuk  MG, Rychlik  K, Shoup  M, Aranha  GV.  Preoperative liver function tests and hemoglobin will predict complications following pancreaticoduodenectomy. J Gastrointest Surg. 2008;12(11):1822-1829.PubMedArticle
Gnerlich  JL, Luka  SR, Deshpande  AD,  et al.  Microscopic margins and patterns of treatment failure in resected pancreatic adenocarcinoma. Arch Surg. 2012;147(8):753-760.PubMedArticle
Michalski  CW, Kleeff  J, Wente  MN, Diener  MK, Büchler  MW, Friess  H.  Systematic review and meta-analysis of standard and extended lymphadenectomy in pancreaticoduodenectomy for pancreatic cancer. Br J Surg. 2007;94(3):265-273.PubMedArticle
Teh  SH, Diggs  BS, Deveney  CW, Sheppard  BC.  Patient and hospital characteristics on the variance of perioperative outcomes for pancreatic resection in the United States: a plea for outcome-based and not volume-based referral guidelines. Arch Surg. 2009;144(8):713-721.PubMedArticle