Secondary Logo

Journal Logo

Pancreaticojejunostomy versus pancreaticogastrostomy after pancreaticoduodenectomy

A review article and meta-analysis of randomized controlled trials

Shahzad, Noman; Chawla, Tabish U; Begum, Saleema; Shaikh, Fareed A

doi: 10.4103/ijssr.ijssr_8_18
Meta-analysis of Randomized Controlled Trials
Free

Pancreatic fistula is one of the most feared complications after pancreaticoduodenectomy (PD). Results of randomized controlled trials comparing pancreaticogastric anastomosis with pancreaticojejunal anastomosis are not consistent. Furthermore, soft texture of pancreas is an established risk factor for postoperative pancreatic fistula (POPF). There has been no meta-analysis to date to compare pancreaticogastrostomy versus pancreaticojejunostomy in patients with intraoperative soft texture of pancreas. Hence, our primary objective was to determine the role of pancreaticogastrostomy compared to pancreaticojejunostomy after PD in prevention of POPF, especially in patients with soft pancreatic texture. We conducted meta-analysis of randomized controlled trials that had compared pancreaticojejunal anastomosis with pancreaticogastric anastomosis after PD, and pancreatic fistula was among the outcome variables. Ten randomized controlled trials were included in the meta-analysis which comprised of a total of 1629 patients, of which 803 underwent pancreaticojejunostomy, whereas 826 were in the pancreaticogastrostomy group. There was no difference in clinically relevant POPF (CR-POPF) rate in pancreaticojejunostomy versus pancreaticogastrostomy (19.8% vs. 12.8%, P = 0.09) group. POPF rate in patients with soft pancreas was significantly more in pancreaticojejunostomy group as compared to pancreaticogastrostomy group (25.4% vs. 17.3%, odds ratio = 1.71, 95% confidence interval = 1.15–2.53, P = 0.008). Although there is no difference in pancreaticogastrostomy as compared to pancreaticojejunostomy after PD to prevent CR POPF, in a subgroup of high-risk patients with soft pancreatic texture pancreaticogastrostomy has favorable results.

Section of General Surgery, Department of Surgery, Aga Khan University and Hospital, Karachi, Pakistan

Section of General Surgery, Department of Surgery, Aga Khan University and Hospital, Karachi, Pakistan

Section of General Surgery, Department of Surgery, Aga Khan University and Hospital, Karachi, Pakistan

Section of General Surgery, Department of Surgery, Aga Khan University and Hospital, Karachi, Pakistan

Address for correspondence:Noman Shahzad, Section of General Surgery, Department of Surgery, Aga Khan University and Hospital, Karachi, Pakistan drns01@hotmail.com

Back to Top | Article Outline

Introduction

Pancreaticoduodenectomy (PD) remains the only curative option for resectable pancreatic head, ampullary, duodenal, and distal common bile duct tumors. Despite improvements in postoperative care and advancements in surgical techniques, morbidity related to this operation remains very high. According to recent report by Tremblay St-Germain et al. up to 74% of patients suffer from at least one complication related to this complex surgical procedure. 1 Leakage of pancreatic enzymes leading to either formation of abdominal collection or pancreatic fistula is one of the most feared complications. The incidence of postoperative pancreatic fistula (POPF) after PD is reported to be from 11% to 47.7% in various reports. 2 , 3 This wide variation in the occurrence of POPF is partly due to variability in definition of fistula, particularly in older studies. For standardization and uniform reporting of POPF to allow comparison across studies, international study group on pancreatic fistula (ISGPF) in 2005 agreed upon an objective and internationally acceptable definition. 4 According to this definition, POPF was labeled if there was drain output of any measurable volume of fluid on or after postoperative day 3 with amylase content >3 times the serum amylase activity. It was further categorized into three grades (Grades A, B, and C) according to clinical impact on patient's hospital course. Later, in 2016, an international study group in pancreatic surgery (ISGPS) updated the definition and essentially reclassified Grade A pancreatic fistula as a biochemical leak and was no longer referred to as true pancreatic fistula, whereas Grades B and C were grouped as clinically relevant POPF (CR-POPF). 5 In light of this update on pancreatic fistula definition, there is a need to incorporate it in studies reporting on pancreatic fistula.

Furthermore, due to high morbidity and cost related to pancreatic fistula, 6 multiple interventions have been investigated to prevent this complication. 7 These include pharmacological interventions such as role of perioperative octreotides administration, adjuncts to surgical anastomosis such as stenting of anastomosis or use of sealants, surgical techniques, and site of pancreaticoenteric anastomosis. Results of randomized controlled trials comparing pancreaticogastric anastomosis with pancreaticojejunal anastomosis are not consistent, with some trials showing significantly less incidence of POPF in patients who underwent pancreaticogastrostomy, 8 , 9 while others failing to detect any significant difference. 10 Pooling of data in reported meta-analysis is also criticized due to variations in inclusion criteria and statistical methodology hence reporting conflicting results. 11 , 12 A recent meta-analysis by Crippa et al. compared various techniques of making anastomosis as well along with pancreaticogastrostomy versus pancreaticojejunostomy. 13 However, they used random effect model to estimate their desired parameters; hence, results were not significant.

In addition to postoperative care and surgical technique, certain patient- and disease-related factors predispose someone to the high risk of POPF development. 14 Soft texture of pancreas is an established risk factor for POPF. 15 There has been no meta-analysis to date to compare pancreaticogastrostomy versus pancreaticojejunostomy in patients with intraoperative soft texture of pancreas.

Objectives

Primary objective

  • To determine the role of pancreaticogastrostomy compared to pancreaticojejunostomy after PD in prevention of POPF
  • To determine the role of pancreaticogastrostomy compared to pancreaticojejunostomy after PD in prevention of POPF in patients with the soft texture of pancreas.

Secondary objective

The secondary objectives of the study were to determine the role of pancreaticogastrostomy compared to pancreaticojejunostomy after PD in prevention of postoperative complications such as biliary fistula, delayed gastric emptying (DGE), hemorrhage, abdominal collections, and mortality.

Back to Top | Article Outline

Methodology

The study protocol was registered at PROSPERO, registration number “CRD42017069361.” Meta-analysis was conducted and reported in accordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses. 16

Eligibility criteria

Inclusion criteria

We conducted a meta-analysis of randomized controlled trials that had compared pancreaticojejunal anastomosis with pancreaticogastric anastomosis after PD, and pancreatic fistula was among the outcome variables. We included all the trials reported in English language till May 31, 2017. There was no restriction to any gender, age, or indication of surgery.

Exclusion criteria

Nonrandomized trials, cohort, case–control, cross-sectional studies were excluded from the study.

Outcomes of interest

Primary outcome measure: Postoperative pancreatic fistula

The primary outcome of interest was the occurrence of POPF. Trials conducted before 2005 had labeled POPF according to definitions used at their centers. After standardization of definition POPF in 2005 by ISGPF, 4 most of reported trials looking at POPF as outcome have used this definition. After 2016 modification by ISGPS to reclassify Grade A pancreatic fistula as biochemical leak rather than true pancreatic fistula, and Grade B and C as CR-POPF, 5 recent trials have incorporated this definition. For purpose of our meta-analysis, we divided our outcome into two categories. One was overall occurrence of POPF and the other one was CR-POPF. For overall POPF, we included all the trials irrespective of their definitions of POPF and pooled the data. While for CR-POPF, we included only those trials which had used ISGPF definition to label POPF and we extracted a number of patients who developed Grade B and Grade C POPF only to include in meta-analysis.

We also performed subgroup analysis of patients who were high risk for developing POPF. For this purpose, we extracted a number of patients with soft texture of pancreas in each group along with their fistula rates. The trials which did not have this information, we requested corresponding authors through E-mail to provide relevant data.

Secondary outcome measures

  • Secondary outcome measures included biliary fistula, DGE, postpancreatectomy hemorrhage, intra-abdominal collections, and mortality
  • Definitions of these outcomes varied across trials, and we relied on primary study definitions to identify positive outcome
  • Mortality was taken as death during index hospitalization or within 30 days of operation wherever reported.

Studies selection process

Two reviewers carried out independent comprehensive systematic literature search in PubMed, Google Scholar, and Cochrane Library. Search terms were selected to identify patient population and intervention which were as follows:

Patient population of interest was those patients undergoing PD for both malignant and benign conditions. We identified “PD,” “Whipple,” and “Pylorus Preserving PD” as terms to look for studies of our interest.

To identify articles which compared pancreaticogastrostomy with pancreaticojejunostomy, terms “pancreaticogastrostomy OR pancreatogastrostomy OR pancreaticogastr* OR pancreatogastr*” and “pancreaticojejunostomy OR pancreaticojejunostomy OR pancreaticojejun* OR pancreatojejun*” were used.

These two groups of terms were used to search for patient population and intervention of interest. All the articles that came out after combined search of terms in included databases were considered for inclusion in meta-analysis. Duplicates in identified articles through different database searches were identified and excluded. Relevant articles were identified through initial screening of titles. Further scrutiny was done by reading abstracts, and final inclusion was decided after full manuscript reviews. References of included articles were also searched to identify further relevant articles. Final inclusion into meta-analysis was done by consensus of both reviewers. In case of conflict, opinion of the third reviewer was sought.

Data extraction

After identification of studies to be included in the meta-analysis, data were extracted from each article by two authors independently. Same consensus procedure was followed for any conflict as described early for inclusion of studies. Data were extracted on standard data collection sheet. It included variables related to study details such as first author name, year of publication, settings of trial, and country. Patients' demographic information such as gender and age and disease-related information such as underlying pathology were also collected. Data regarding primary and secondary outcomes measures were extracted. In case of missing data, corresponding authors were contacted through E-mail to provide relevant data.

Statistical methods

All statistical analyses were carried out using Review Manager Version 5.3 (The Cochrane Collaboration, Oxford, United Kingdom). 17 Odds ratios (ORs) with 95% confidence intervals (CIs) were used to analyze dichotomous outcome variables. Variability among studies included in the specific analysis was assessed using I 2values. 18 Any value above 50% indicated substantial heterogeneity in which case random-effect model was used, while in cases, where I 2value was ≤50%, a fixed-effect model was used.

Risk of bias assessment

Possibility of reporting bias was assessed by making funnel plot for the primary outcome and was evaluated by visual inspection.

Back to Top | Article Outline

Results

Initial search strategy retrieved 1128 studies, of which 332 studies were duplicates. Screening of remaining 796 studies by titles and then by abstracts resulted in 15 studies to be reviewed as a full manuscript. Five articles were excluded at this stage as they did not either have intervention of interest or the outcome of interest. Details are as given in Figure 1. Ten randomized controlled trials included in the meta-analysis comprised a total of 1629 patients, of which 803 underwent pancreaticojejunostomy, whereas 826 were in the pancreaticogastrostomy group.

Figure 1

Figure 1

Characteristics of included studies

All ten studies included in meta-analysis were randomized controlled trials. Earliest reported trial was from 1995. A number of trial participants ranged from 90 to 329. Details of study characteristics are as shown in Table 1and Table 2.{Table 1}{Table 2}

Publication bias

Funnel plot as shown in Figure 2was made to assess the risk of biased reporting. All the dots were within 95% confidence limits, and dots are distributed evenly about the vertical line, indicating low risk of biased reporting.

Figure 2

Figure 2

Results of meta-analyses

Postoperative pancreatic fistula

All the included studies reported rate of POPF. Overall POPF rate was 21.85% in pancreaticojejunostomy group, whereas it was 16.83% in pancreaticogastrostomy group. Meta-analysis using fixed effect model showed significant difference in favor of pancreaticogastrostomy (OR = 1.39, 95% CI = 0.32–2.45, P = 0.009, I 2= 10%). Results are as shown in Figure 3.

Figure 3

Figure 3

Clinically relevant postoperative pancreatic fistula

Seven studies used definition by ISGPF for pancreatic fistula. Information was extracted regarding CR POPF as Grade B and Grade C pancreatic fistula. CR-POPF rate was 19.28% in pancreaticojejunostomy group, whereas it was 12.8% in pancreaticogastrostomy group. Meta-analysis using random-effect model showed insignificant trend in favor of pancreaticogastrostomy (OR = 1.64, 95% CI = 0.92–2.92, P = 0.09, I 2= 61%). Results are as shown in Figure 4.

Figure 4

Figure 4

Postoperative pancreatic fistula in patients with soft pancreatic texture

Two studies reported rate of POPF separately for patients who had soft pancreatic texture. Authors of other studies were contacted for details of POPF in this subgroup and response was received from three authors. Data were included in the analysis. Overall POPF rate in patients with soft pancreas was 25.4% in pancreaticojejunostomy group, while it was 17.3% in pancreaticogastrostomy group. Meta-analysis using fixed-effect model showed significant difference in favor of pancreaticogastrostomy (OR = 1.71, 95% CI = 1.15–2.53, P = 0.008, I 2= 27%). Results are as shown in Figure 5.

Figure 5

Figure 5

Biliary fistula

Five studies reported the incidence of biliary fistula formation. Overall biliary fistula rate was 6.7% in pancreaticojejunostomy group, whereas it was 2.6% in pancreaticogastrostomy group. Meta-analysis using fixed-effect model showed significant difference in favor of pancreaticogastrostomy (OR = 2.52, 95% CI = 1.14–5.67, P = 0.02, I 2= 38%). Results are as shown in Figure 6.

Figure 6

Figure 6

Delayed gastric emptying

Eight studies reported rate of DGE. Rate of DGE was 19.7% in pancreaticojejunostomy group, whereas it was 21.7% in pancreaticogastrostomy group. Meta-analysis using random-effect model showed insignificant trend in favor of pancreaticojejunostomy (OR = 0.93, 95% CI = 0.59–1.46, P = 0.75, I 2= 53%). Results are as shown in Figure 7.

Figure 7

Figure 7

Intra-abdominal collection

Incidence of intra-abdominal collection was reported by six studies. It was 13.44% in pancreaticojejunostomy group, whereas 10.29% in pancreaticogastrostomy group. Meta-analysis using fixed-effect model showed insignificant trend in favor of pancreaticogastrostomy (OR = 1.41, 95% CI = 0.98–2.03, P = 0.07, I 2= 49%). Results are as shown in Figure 8.

Figure 8

Figure 8

Postpancreatectomy hemorrhage

Seven studies reported rate of postpancreatectomy hemorrhage. The rate was 8.16% in pancreaticojejunostomy group, whereas it was 13.46% in pancreaticogastrostomy group. Meta-analysis using fixed-effect model showed significant difference in favor of pancreaticojejunostomy (OR = 0.59, 95% CI = 0.4–0.85, P = 0.005, I 2= 0%). Results are as shown in Figure 9.

Figure 9

Figure 9

Postoperative mortality

Postoperative mortality was reported in eight studies. It was mortality during index hospitalization in most of the included studies except Topal et al., who reported 60-day mortality. Overall mortality rate was 4.6% in pancreaticojejunostomy group, whereas it was 5.01% in pancreaticogastrostomy group. Meta-analysis using fixed effect model showed insignificant trend in favor of pancreaticojejunostomy (OR = 0.96, 95% CI = 0.58–1.57, P = 0.86, I 2= 0%). Results are as shown in Figure 10.

Figure 10

Figure 10

Back to Top | Article Outline

Discussion

This meta-analysis has shown no difference in the occurrence of CR-POPF after pancreaticogastrostomy as compared to pancreaticojejunostomy. This is similar to the results from meta-analysis by Crippa et al., 13 but contradictory to other meta-analyses reported before that which showed results in favor of pancreaticogastrostomy. A recent multicentric randomized controlled trial by Keck et al. 24 is the largest trial reported as yet including a total of 171 patients. Results of this trial showed that pancreaticogastrostomy is comparable to pancreaticojejunostomy so far as POPF is concerned. Inclusion of this trial in meta-analysis has contributed 23.4% weightage to the results. Although overall results are insignificant, the trend is in favor of pancreaticogastrostomy.

Contrary to that, meta-analysis of occurrence of POPF in patients with soft pancreatic texture showed significantly low risk after pancreaticogastrostomy as compared to pancreaticojejunostomy. This is first meta-analysis to report on a comparison of pancreaticogastrostomy with pancreaticojejunostomy in high-risk patients' population. It has been proposed that lack of enterokinase and acidic environment in stomach inactivates pancreatic enzymes, which along with good blood supply of stomach may have a role to play in reducing the risk of anastomotic leak. 25 While potential of anastomotic leak is reduced by pancreaticogastrostomy, especially in patients with soft pancreas, long-term exocrine and endocrine functions are compromised more in these patients as compared to those who underwent pancreaticojejunostomy. 26 Furthermore, the risk of digestive tract bleeding is also more after pancreaticogastrostomy, which is reported by other meta-analyses as well. 27

Despite uniformity of definition of POPF for comparison across studies, surgical technique of doing pancreaticoenteric anastomosis remains variable and largely depends on surgeons' training and experience of doing one type of anastomosis. Uncommon variations of techniques were also compared in some trials. For example, El Nakeeb et al. 19 compared isolated Roux loop pancreaticojejunostomy to pancreaticogastrostomy, while Fernández-Cruz et al. 20 compared pancreaticogastrostomy with a gastric partition to pancreaticojejunostomy. These individual trial variations can impact the results.

Back to Top | Article Outline

Conclusion

Although there is no difference in pancreaticogastrostomy as compared to pancreaticojejunostomy after PD to prevent CR POPF, in a subgroup of high-risk patients with soft pancreatic texture pancreaticogastrostomy has favorable results. Digestive tract bleeding is more common after pancreaticogastrostomy. There is no difference in DGE, intra-abdominal collections, and mortality.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

1. Tremblay St-Germain A, Devitt KS, Kagedan DJ, Barretto B, Tung S, Gallinger S, et al The impact of a clinical pathway on patient postoperative recovery following pancreaticoduodenectomy HPB (Oxford). 2017;19:799–807
2. Yeo CJ, Cameron JL, Maher MM, Sauter PK, Zahurak ML, Talamini MA, et al Aprospective randomized trial of pancreaticogastrostomy versus pancreaticojejunostomy after pancreaticoduodenectomy Ann Surg. 1995;222:580–8
3. Aumont O, Dupré A, Abjean A, Pereira B, Veziant J, Le Roy B, et al Does intraoperative closed-suction drainage influence the rate of pancreatic fistula after pancreaticoduodenectomy? BMC Surg. 2017;17:58
4. Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C, Izbicki J, et al Postoperative pancreatic fistula: An international study group (ISGPF) definition Surgery. 2005;138:8–13
5. Bassi C, Marchegiani G, Dervenis C, Sarr M, Abu Hilal M, Adham M, et al The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 years after Surgery. 2017;161:584–91
6. Williamsson C, Ansari D, Andersson R, Tingstedt B. Postoperative pancreatic fistula-impact on outcome, hospital cost and effects of centralization HPB (Oxford). 2017;19:436–42
7. Lai EC, Lau SH, Lau WY. Measures to prevent pancreatic fistula after pancreatoduodenectomy: A comprehensive review Arch Surg. 2009;144:1074–80
8. Topal B, Fieuws S, Aerts R, Weerts J, Feryn T, Roeyen G, et al Pancreaticojejunostomy versus pancreaticogastrostomy reconstruction after pancreaticoduodenectomy for pancreatic or periampullary tumours: A multicentre randomised trial Lancet Oncol. 2013;14:655–62
9. Figueras J, Sabater L, Planellas P, Muñoz-Forner E, Lopez-Ben S, Falgueras L, et al Randomized clinical trial of pancreaticogastrostomy versus pancreaticojejunostomy on the rate and severity of pancreatic fistula after pancreaticoduodenectomy Br J Surg. 2013;100:1597–605
10. Grendar J, Ouellet JF, Sutherland FR, Bathe OF, Ball CG, Dixon E, et al In search of the best reconstructive technique after pancreaticoduodenectomy: Pancreaticojejunostomy versus pancreaticogastrostomy Can J Surg. 2015;58:154–9
11. Qin H, Luo L, Zhu Z, Huang J. Pancreaticogastrostomy has advantages over pancreaticojejunostomy on pancreatic fistula after pancreaticoduodenectomy.A meta-analysis of randomized controlled trials Int J Surg. 2016;36:18–24
12. Zhang X, Ma L, Gao X, Bao H, Liu P, Aziz A, et al Pancreaticogastrostomy versus pancreaticojejunostomy reconstruction after pancreaticoduodenectomy: A meta-analysis of randomized controlled trials Surg Today. 2015;45:585–94
13. Crippa S, Cirocchi R, Randolph J, Partelli S, Belfiori G, Piccioli A, et al Pancreaticojejunostomy is comparable to pancreaticogastrostomy after pancreaticoduodenectomy: An updated meta-analysis of randomized controlled trials Langenbecks Arch Surg. 2016;401:427–37
14. Søreide K, Labori KJ. Risk factors and preventive strategies for post-operative pancreatic fistula after pancreatic surgery: A comprehensive review Scand J Gastroenterol. 2016;51:1147–54
15. Yang MW, Deng Y, Huang T, Zhang LD. Clinical study on the relationship between pancreatic fistula and the degree of pancreatic fibrosis after pancreatic and duodenal resection Zhonghua Wai Ke Za Zhi. 2017;55:373–7
16. Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement Int J Surg. 2010;8:336–41
17. Copenhagen , The Nordic Cochrane Centre: The Cochrane Collaboration. Review Manager (RevMan).Ver . 5;: Review Manager (RevMan) Ver 53 Copenhagen: The Nordic Cochrane Centre: The Cochrane Collaboration; 2013
18. Higgins J, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses [journal article as teaching resource, deposited by John Flynn] Br Med J. 2003;327:557–60
19. El Nakeeb A, Hamdy E, Sultan AM, Salah T, Askr W, Ezzat H, et al Isolated roux loop pancreaticojejunostomy versus pancreaticogastrostomy after pancreaticoduodenectomy: A prospective randomized study HPB (Oxford). 2014;16:713–22
20. Fernández-Cruz L, Cosa R, Blanco L, López-Boado MA, Astudillo E. Pancreatogastrostomy with gastric partition after pylorus-preserving pancreatoduodenectomy versus conventional pancreatojejunostomy: A prospective randomized study Ann Surg. 2008;248:930–8
21. Wellner UF, Sick O, Olschewski M, Adam U, Hopt UT, Keck T, et al Randomized controlled single-center trial comparing pancreatogastrostomy versus pancreaticojejunostomy after partial pancreatoduodenectomy J Gastrointest Surg. 2012;16:1686–95
22. Duffas JP, Suc B, Msika S, Fourtanier G, Muscari F, Hay JM, et al Acontrolled randomized multicenter trial of pancreatogastrostomy or pancreatojejunostomy after pancreatoduodenectomy Am J Surg. 2005;189:720–9
23. Bassi C, Falconi M, Molinari E, Salvia R, Butturini G, Sartori N, et al Reconstruction by pancreaticojejunostomy versus pancreaticogastrostomy following pancreatectomy: Results of a comparative study Ann Surg. 2005;242:767–71
24. Keck T, Wellner UF, Bahra M, Klein F, Sick O, Niedergethmann M, et al Pancreatogastrostomy versus pancreatojejunostomy for RECOnstruction after PANCreatoduodenectomy (RECOPANC, DRKS 00000767): Perioperative and long-term results of a multicenter randomized controlled trial Ann Surg. 2016;263:440–9
25. Machado NO. Pancreatic fistula after pancreatectomy: Definitions, risk factors, preventive measures, and management-review Int J Surg Oncol. 2012;2012:602478
26. Roeyen G, Jansen M, Ruyssinck L, Chapelle T, Vanlander A, Bracke B, et al Pancreatic exocrine insufficiency after pancreaticoduodenectomy is more prevalent with pancreaticogastrostomy than with pancreaticojejunostomy.A retrospective multicentre observational cohort study HPB (Oxford). 2016;18:1017–22
27. Clerveus M, Morandeira-Rivas A, Picazo-Yeste J, Moreno-Sanz C. Pancreaticogastrostomy versus pancreaticojejunostomy after pancreaticoduodenectomy: A systematic review and meta-analysis of randomized controlled trials J Gastrointest Surg. 2014;18:1693–704
28. . .;:
    Keywords:

    Pancreatic fistula; pancreaticoduodenectomy; pancreaticogastrostomy; pancreaticojejunostomy

    © 2018 President and Fellows of Harvard College