Skip to main content
NCBI home page
HPB : The Official Journal of the International Hepato Pancreato Biliary Association logoLink to HPB : The Official Journal of the International Hepato Pancreato Biliary Association
. 2014 Jul 7;17(2):113–122. doi: 10.1111/hpb.12299

The impact of pancreaticojejunostomy versus pancreaticogastrostomy reconstruction on pancreatic fistula after pancreaticoduodenectomy: meta-analysis of randomized controlled trials

Julie Hallet 1,2, Francis S W Zih 1, Raymond G Deobald 1,2, Adena S Scheer 1,2, Calvin H L Law 1,2, Natalie G Coburn 1,2, Paul J Karanicolas 1,2,✉
PMCID: PMC4299385  PMID: 25040921

Abstract

Background

Pancreatic fistula (PF) remains a common source of morbidity following pancreaticoduodenectomy (PD). Despite numerous studies, the optimal method of pancreatic remnant reconstruction is controversial. This study examines the hypothesis that pancreaticogastrostomy (PG) is associated with a lower risk for PF after PD compared with pancreaticojejunostomy (PJ).

Methods

Five electronic databases and the grey literature were searched for randomized controlled trials (RCTs) comparing PJ and PG after PD. Two reviewers independently selected studies, extracted data and assessed methodology. The primary outcome was the occurrence of PF of International Study Group on Pancreatic Fistula (ISGPF) Grade B or C.

Results

Four RCTs including 676 patients were included. Pancreaticogastrostomy reduced the risk for PF [relative risk (RR) 0.41, 95% confidence interval (CI) 0.21–0.62] without any difference between high- and low-risk patients. Absolute risk reduction for PF was 4% (95% CI 2.4–5.6) in low-risk patients compared with 10% (95% CI 6.5–14.8) in high-risk patients undergoing PG rather than PJ. The strength of evidence for PF outcome was moderate according to the GRADE classification.

Conclusions

Reconstruction by PG decreases the rate of PF in comparison with PJ. Surgeons should consider reconstructing the pancreatic remnant following PD with PG, particularly in patients at high risk for PF.

Introduction

Advances in pancreatic surgery techniques and perioperative care have led to reduced mortality rates for pancreaticoduodenectomy (PD) in high-volume expert centres.1,2 However, morbidity after pancreatic resection remains high, with 30–60% of patients experiencing complications following surgery, mainly as a result of leak and subsequent fistula from the pancreatic anastomosis.2,3

Pancreatic reconstruction is particularly demanding; a variety of methods and techniques have been proposed to maintain the continuity of the anastomosis and diminish rates of leak.4–7 The abundance of literature on this issue reflects the ongoing controversy regarding the optimal method of pancreatic anastomosis. The conventional anastomosis described for this operation is pancreaticojejunostomy (PJ).8 Pancreaticogastrostomy (PG) has been described and studied as an alternative to jejunal anastomosis in both observational studies and randomized controlled trials (RCTs) with inconsistent results.9–11

Given the ongoing frequency of pancreatic leak or fistula and the major morbidity associated with this complication, robust evidence is needed to determine which reconstruction technique yields better outcomes. Five systematic reviews and meta-analyses have attempted to summarize the impact of PG compared with PJ on the occurrence of pancreatic fistula (PF); however, most included small numbers of RCTs among large numbers of observational studies, suffered from important methodological limitations, lacked a consensus definition of PF and sensitivity analyses, and failed to identify a significant benefit of either technique.10,12–14 More recently, two larger RCTs examining this important question have been published.9,15 The present systematic review and meta-analysis of RCTs was undertaken to examine the impact of PG compared with that of PJ on PF in patients undergoing PD, and to address methodological issues arising from previous meta-analyses.

Materials and methods

This review was registered in PROSPERO (2013: CRD42013005288).16

Search strategy

MEDLINE (1966 to August 2013), EMBASE (1974 to August 2013), the Cochrane Central Register for Controlled Trials, Web of Knowledge, and the Scopus database (1966 to August 2013) were systematically searched, with the help of an information specialist, to identify potential RCTs, without language or other limitations. The following search terms were used: (i) pancreatic dis$, pancreatic neo$, pancreas cancer, chronic pancreatitis, and (ii) pancreatectomy, pancreatic resection, pancreaticojej$, pancreaticogas$, pancreas reconstruction, and pancreas anast$. A standardized filter was applied for RCTs. The grey literature (informally published material not indexed in formal search engines) was searched for unpublished results using the OpenSIGLE database, Trip database, Google Scholar, and the database of registered trials (http://www.clinicaltrials.gov). Bibliographies of all included studies and previous narrative or systematic reviews were also reviewed for relevant publications.

Two authors (FSWZ and RGD) independently selected studies, extracted data and assessed the risk for bias. Disagreements were resolved by consensus or by a third party (JH). A single reviewer (JH) assessed the references from the grey literature.

Study selection

Explicit eligibility criteria allowed the inclusion of RCTs reporting the effects on PF of PG compared with PJ PD. Studies including at least 10 adults (aged ≥18 years) submitted to PD for benign or malignant disease, and defining PF according to the International Study Group on Pancreatic Fistula (ISGPF) definitions were eligible.17 Studies that examined PD for trauma or acute pancreatitis were excluded. Studies that included patients who did not meet the present inclusion criteria were excluded if it was not possible to distinguish those patients from the larger population. In the event of duplicate publications, the most relevant and most informative study was included.

Data abstraction

A standardized data extraction form was developed and pilot-tested following the recommendations of the Cochrane Effective Practice and Organization of Care Review Group.18 The following patient characteristics were collected: age; gender, and indication for surgery. The corresponding author for each study was contacted to obtain additional details about missing or incomplete data when this was deemed necessary. Pancreas-associated risk was categorized based on baseline risk for PF when possible, as reported by the authors in the original manuscript as: (i) high risk for PF (soft gland or small pancreatic duct), or (ii) low risk for PF (hard gland or large pancreatic duct).

Outcome measures

The primary outcome was the occurrence of clinically significant PF defined as Grade B or C fistula based on the ISGPF classification.17 Secondary outcomes included operating time (min), estimated blood loss (ml), bile leak, delayed gastric emptying (DGE), postoperative bleeding (divided into intraluminal and extraluminal bleeding), 30-day postoperative major morbidity (Clavien–Dindo Grades III and IV) and mortality, and length of stay (LoS) (days).19–21

Risk for bias assessment

The methodological quality of included studies was evaluated using a checklist of key methodological components derived from the CONSORT (consolidated standards of reporting trials) statement rather than by using a summary score.22–25 The GRADE (grading of recommendations assessment, development and evaluation) system was used to present a summary of findings and rate the overall strength of evidence.26 The robustness of included RCTs that reported a significant difference in PF favouring PG was assessed by computing the Fragility Index (FI). This index assesses the minimum number of patients that are required to switch from non-event (no PF) to event (PF) in the group with fewer events (PG) in order to reverse statistical significance.27 Thus, it evaluates how many patients in the PG group would have needed to experience a PF for the difference between PG and PJ to become non-significant.

Statistical analysis

Descriptive statistics were presented as means and standard deviations (SDs) for continuous variables, and as proportions with 95% confidence intervals (95% CIs) for dichotomous variables. When studies presented medians and ranges, the mean and SD were estimated using the method of Hozo et al.28 Meta-analysis was conducted using Review Manager (RevMan) Version 5.2.5 (Cochrane Collaboration, Copenhagen, Denmark) for each outcome with data in two or more studies. The data were pooled for each outcome using random-effects models. The relative risk (RR) with 95% CI was calculated for dichotomous outcomes. Absolute risk reduction (ARR) in PF with PG was computed using the pooled RR for risk reduction and PF rate with PJ for baseline risk. The standard mean difference (95% CI) was used for continuous variables.18 The I2 statistic was used to assess the extent of heterogeneity.29 The following a priori hypotheses were generated to explain heterogeneity in results: (i) study design (single-centre versus multicentre studies); (ii) population (high- versus low-risk pancreas, and malignant versus benign indications for surgery); (iii) intervention (duct-to-mucosa PJ versus a ‘fish-mouth’ PJ technique), and (iv) intervention (pancreatic stent versus no stent). Publication bias was assessed visually by creating a funnel plot for each outcome for which data were available from five or more trials.18

The review was conducted and results reported in accordance with the PRISMA (preferred reporting items for systematic reviews and meta-analyses) statement.30

Results

Systematic search

Four RCTs enrolling a total of 676 patients were included (Fig. 1).9,15,31,32

Figure 1.

Figure 1

Open in a new tab

Study selection conducted in line with PRISMA (preferred reporting items for systematic reviews and meta-analyses) guidelines

Searches of the grey literature and of the references of included studies did not identify additional citations. One additional reference was identified after the initial electronic search in the table of contents of a surgical journal, and included in the analysis.15

Description of included studies

All studies were published in English between 2008 and 2013 and included between 108 and 329 patients (Table 1). Study techniques and postoperative management protocols are presented in Table 2.

Table 1.

Populations and clinical characteristics of the included studies

Study Study period Population characteristics
 High-risk pancreas n (%)
Sample, n Pancreatic cancer n (%) Age, years old, mean (SE or range)
Fernández-Cruz et al.31 2005–2007 108 PG: 26 (49%) PG: 63 (13) PG: 24 (45%)
2008 PJ: 28 (51%) PJ: 63 (14) PJ: 25 (45%)
Spain
Wellner et al.32 2006–2011 116 PG: 26 (44%) PG: 67 (34–84) PG: 35 (59%)
2012 PJ: 30 (53%) PJ: 64 (23–81) PJ: 29 (51%)
Germany
Topal et al.9 2009–2012 329 PG: 98 (61%) PG: 67 (61–73) PG: 98 (60%)
2013 PJ: 107 (64%) PJ: 66 (59–75) PJ: 102 (61%)
Belgium
Figueras et al.15 2008–2012 123 PG: 55 (85%) PG: 67 (35–80) PG: 34 (52%)
2013 PJ: 49 (84%) J: 66 (42–80) PJ: 33 (57%)
Spain
Open in a new tab

PG, pancreatogastrostomy;

PJ, pancreatojejunostomy;

SE, standard error.

Table 2.

Surgical management in the included studies

Study Surgical technique
 Postoperative management
Surgeons, n PJ technique PG technique Pylorus-preserving n (%) Use of stent Use of drain Prophylactic octreotide
Fernández-Cruz et al.31 2008 NR 2 layers Duct-to-mucosa End-to-side 2 layers Duct-to-mucosa To bottom of gastric partition PG: 53 (100%) PJ: 55 (100%) Yes, internal Yes No
Wellner et al.32 2012 3 2 layers Duct-to-mucosa 2 layers Dunking To posterior gastric wall PG: 52 (88%) PJ: 55 (96%) Yes, external (only for PJ) Yes Yes, selectivea
Topal et al.9 2013 NR Layers based on surgeon decision Dunking End-to-side Layers based on surgeon decision Dunking To posterior gastric wall PG: 98 (60%) PJ: 102 (61%) No Yes Yes, all
Figueras et al.15 2013 3 2 layers Duct-to-mucosa 2 layers Dunking To posterior gastric wall PG: 30 (46%) PJ: 28 (48%) NR Yes No
Open in a new tab
a

Information from referenced publication.33

PG, pancreatogastrostomy;

PJ, pancreatojejunostomy;

NR, not reported.

Key methodological elements are summarized in Table 3. The FIs of the studies reporting a significant benefit were 2 (n = 108),31 6 (n = 329),9 and 5 (n = 123).15

Table 3.

Assessment of methodology

Study Setting Timing of randomization Stratification of randomization Allocation concealment Blinding Standardization of technique Pancreatic fistula measure
Fernández-Cruz et al.31 2008 Single centre Before surgery No No NR Yes ISGPF
Wellner et al.32 2012 Single centre During surgery, before reconstruction No No NR Yes ISGPF
Topal et al.9 2013 Multiple centres During surgery, before reconstruction Yes (pancreas duct size) Yes No No ISGPF
Figueras et al.15 2013 Multiple centres During surgery, before reconstruction Yes (diagnosis and centre) Yes NR Yes ISGPF
Open in a new tab

NR, not reported;

ISGPF, International Study Group on Pancreatic Fistula.

Primary outcome: pancreatic fistula

Rates of PF varied from 18% (n = 10/55) to 33% (n = 19/58) in the PJ control groups, and from 11% (n = 12/108) to 21% (n = 26/123) in entire cohorts. Patients undergoing PG had a significantly lower risk for PF than patients who underwent PJ (RR 0.41, 95% CI 0.27–0.62) (Fig. 2). The funnel plot did not show convincing asymmetry. The strength of evidence supporting risk reduction in PF with PG was moderate (Table 4).

Figure 2.

Figure 2

Open in a new tab

Comparison of the occurrence of pancreatic fistula after pancreaticogastrostomy (PG) with that after pancreaticojejunostomy (PJ). IV, inverse variance; 95% CI, 95% confidence interval

Table 4.

GRADE summary of findings for the effect of pancreaticogastrostomy versus pancreaticojejunostomy for reconstruction after pancreaticoduodenectomy

Outcome Participants, n (studies, n) Illustrative comparative risk
 Summary measure (95% CI)a Absolute risk reduction (95% CI)b Quality of evidence (GRADE)c
PG PJ
Clinically significant pancreatic fistula 1133 (4) 8% (28/339) 20% (69/337) RR 0.41 (0.27–0.62) +++ Moderate
Among high-risk pancreas 69 (2) 11% (15/133) 24% (32/131) 10% (6.5–14.8) +++ Moderate
Among low-risk pancreas 180 (2) 5% (4/87) 9% (8/93) 4% (2.4–5.6) ++ Low
Operative time (min) 676 (4) – – SMD 0.01 (−0.22 to 0.25) – ++ Low
Estimated blood loss (ml) 560 (3) – – SMD 0.20 (−0.12 to 0.53) ++ Low
Biliary leak 231 (2) 0.8% (1/118) 6% (7/113) RR 0.18 (0.03–1.07) ++ Low
Delayed gastric emptying 676 (4) 18% (60/339) 13% (45/337) RR 1.23 (0.70–2.16) ++ Low
Postoperative bleeding 676 (4) 12% (41/339) 9% (29/337) RR 1.39 (0.89–2.17) ++ Low
Extraluminal bleedingd 32 (3) 50% (10/20) 92% (11/12) RR 0.65 (0.41–1.04) ++ Low
Intraluminal bleedingd 30 (2) 58% (11/19) 9% (1/11) RR 3.79 (0.87–16.91) ++ Low
Major postoperative morbidity 437 (2) 52% (112/215) 55% (123/222) RR 0.78 (0.40–1.53) ++ Low
Postoperative mortality 676 (4) 2% (8/339) 4% (12/337) RR 0.66 (0.27 −1.6) ++ Low
Length of stay (days) 676 (4) – – SMD −0.31 (−1.03 to 0.42) ++ Low
Open in a new tab
a

Values of <1 favour PG.

b

Values are the expected difference in absolute risk if patients undergo PG instead of PJ.

c

Quality rated from + (very low) to ++++ (very high).

d

Data reported among the subgroup of patients with postoperative bleeding.

PG, pancreaticogastrostomy; PJ, pancreaticojejunostomy; 95% CI, 95% confidence interval; RR, relative risk; SMD, standard mean difference.

In studies that did not report the use of pancreatic stents, the summary estimate was similar (RR 0.46, 95% CI 0.12–1.79) but non-significant and was associated with greater heterogeneity (I2 = 58%). Finally, an analysis of only multicentre trials revealed a significant reduction in PF with PG (RR 0.38, 95% CI 0.23–0.61; I2 = 0%), compared with a non-significant result for single-centre trials (RR 0.46, 95% CI 0.12–1.76; I2 = 56%).

Secondary outcomes

One study used a consensus definition for DGE and one used Clavien–Dindo classifications to report morbidity. No significant difference was observed in secondary outcomes between PG and PJ (Figs 3 and 4). None of the funnel plots for secondary outcomes showed convincing asymmetry. The quality of evidence was low for all secondary outcomes (Table 4).

Figure 3.

Figure 3

Open in a new tab

Comparisons of the occurrences of biliary leak, delayed gastric emptying, postoperative morbidity (Clavien–Dindo Grades III and IV) and postoperative mortality after pancreaticogastrostomy (PG) with occurrences after pancreaticojejunostomy (PJ). M–H, Mantel–Haenszel method; IV, inverse variance; 95% CI, 95% confidence interval

Figure 4.

Figure 4

Open in a new tab

Comparisons of operating time, estimated blood loss, and length of stay after pancreaticogastrostomy (PG) with those after pancreaticojejunostomy (PJ). IV, inverse variance; 95% CI, 95% confidence interval

Discussion

This study systematically reviewed and pooled data from four RCTs investigating the impacts of PG compared with PJ on PF. Based on evidence of moderate quality, PG is associated with a lower occurrence of PF (RR 0.41, 95% CI 0.27–0.62), but no significant differences emerged in biliary leak, DGE, postoperative bleeding, major morbidity, mortality or LoS. When only high- or low-risk pancreas groups were considered, there was no difference in RR. However, when baseline risks for PF were considered, ARR was greater in patients with a high-risk pancreas (10%) compared with those with a low-risk pancreas (4%), which indicates that PG offers more critical benefits in high-risk patients.

The proposed technical and physiological advantages of PG over PJ have been discussed in several studies reporting the technique.34–38 The anastomosis may be facilitated by a thick gastric wall, can rely on an excellent gastric blood supply, and is subject to less tension as a result of the anatomic proximity of the pancreatic remnant to the posterior gastric wall. Lack of enterokinase in the gastric remnant may prevent the activation of pancreatic enzymes, thereby avoiding both damage to the anastomosis itself and the repercussions associated with potential PF. A nasogastric tube may be used to decompress the stomach and relieve tension on the anastomosis immediately after surgery, and can eventually be reinserted to deal with ileus or manage a leak without percutaneous intervention. Finally, PG can be accessed endoscopically for instrumentation if needed, and is located away from major vessels skeletonized during resection, which may theoretically mitigate the risk for vascular damage by proteolytic pancreatic enzymes in the event of PF. Retrospective studies have raised concerns regarding increased intraluminal bleeding and compromised longterm pancreas exocrine function with PG.12 Although this analysis revealed no difference in the overall risk for postoperative bleeding, patterns appeared to differ among those experiencing it, with more intraluminal events with PG and more extraluminal events with PJ. The former occurs early in the postoperative course and is theoretically more easily managed than the latter, which can potentially represent a devastating delayed result of PF. Because of the very small number of patients experiencing bleeding overall, no significant difference could be detected. Additional concerns regarding the longterm outcomes and durability of PG have also been voiced. Although it was not a specific outcome of this review, one RCT assessed exocrine pancreatic function and identified poorer function with PJ.15 Interestingly, some have suggested that PJ anastomotic stricture resulting from PF is associated with pancreatic insufficiency, which may explain this finding.39

A soft pancreas and small pancreatic ducts have been associated repeatedly with significantly higher rates of PF. The trials included in this meta-analysis included pancreases of both low and high risk, resulting in a heterogeneous population in terms of baseline risk. Although the RR did not change when high- or low-risk pancreas subgroups were considered, the ARR was larger for patients with a high-risk pancreas (10% versus 4%) submitted to PG rather than PJ as a result of the considerable differences in baseline risk for PF. This highlights the more pronounced benefits of PG on PF in the context of a high-risk pancreas.

Some potential differences between PG and PJ may not be obvious in the current analysis as a result of the heterogeneous nature of the studied population. Firstly, there are potential negative impacts of PG compared with PJ on postoperative management, mainly arising from the need for nasogastric tube decompression if the stomach fills with pancreatic secretions and the subsequent delay in oral intake, and the routine use of surgical drains. As a result, PG may potentially be associated with a longer LoS. No difference in LoS was observed in this meta-analysis. Mean lengths of stay were quite long (range: 12–16 days), indicating that the studies included were probably conducted before the adoption of ‘fast-track’ protocols, when postoperative nasogastric tubes, surgical drains and delayed oral intake were routine for all patients.40,41 It is possible that with contemporary clinical pathways and enhanced recovery after surgery protocols, patients submitted to PG and not experiencing a leak would have a longer LoS than those with PJ who are fast-tracked through the postoperative course. Nevertheless, the reduction of PF obviously takes priority over a potentially longer LoS. Secondly, despite the reduction in PF, DGE, morbidity and mortality did not differ between patients submitted to PG and PJ, respectively. With regard to DGE, this may also be explained by the heterogeneity of the population, whereby the lower rate of PF in PG balances out a potentially higher occurrence of DGE in this group. Given the smaller ARR in PF associated with the use of PG in a low-risk pancreas, PG may not appear to be best suited to use in this context. However, because of the larger ARR associated with PG in patients with a high-risk pancreas, the benefits of reducing PF outweigh the potential drawbacks of the technique. Hence, a selective approach to the use of PG after PD based on the baseline risk for PF may be advisable in order to maximize the benefits of the technique. New tools that enable the surgeon to appreciate a patient's risk for PF, such as the fistula risk score, are now available and may help in implementing this selective use of PG.42

Among previous meta-analyses, two observed a significant benefit of PG on the occurrence of PF10,12 and three did not.13,14,43 However, with one exception, these previous studies suffered from important methodological limitations, including a search strategy limited to one or a few electronic databases of material published only in English, the absence of systematic methodological assessment, the lack of sensitivity analyses to explore the heterogeneity of results, and the inclusion of a large number of observational studies, along with a few RCTs.12–14,31 A meta-analysis by McKay et al. used sound methodology, but included only one RCT.10 Therefore, the pooling of previous data to examine the impacts of PG versus PJ on PF was limited by the inclusion of data from inherently biased observational studies. The present meta-analysis is the first to include only RCTs and to observe a significant reduction in PF with PG. It also offers insights into the differences in results through an exploration of heterogeneity and subgroup analyses, leading to a detailed practical proposal for the transfer of results into practice in order to maximize the benefits of selecting the use of PG in patients with a high-risk pancreas.

Although meta-analyses offer increased power to detect treatment effects by synthesizing the results of individual studies, they remain limited by the quality of the studies included and the heterogeneity of results. The weaknesses of this study refer to the methodological limitations of the studies included, most of which are related to allocation concealment and blinding, which are known limitations of surgical RCTs.44,45 Nevertheless, the summary of the findings of this study indicates that conclusions on the primary outcome of PF are based on evidence of moderate quality. The FI analysis shows that at least two of the three trials reporting a benefit for PG are robust, with indices of 5 and 6.9,15,31 This index indicates that the results are unlikely to be changed by the transforming of only a few events (of PF) into non-events (of no PF) and thus can be relied upon. Such a finding is closely related to sample size: studies with larger sample sizes are less prone to fragile results in which the significance of the measure of effect can be modified by the transformation of only one or two events into non-events.

The strengths of this systematic review and meta-analysis include a comprehensive, systematic and highly selective search conducted without restriction for language or types of publication, which also considered the grey literature and was performed in duplicate. The studies included were rigorously evaluated using a checklist of key methodological elements drawn from a consensus statement on RCTs, GRADE quality of evidence classification, and the FI to assess the robustness of positive results. The primary outcome chosen was based on a consensus definition, was common, was significant in terms of clinical repercussions, and was important to patients. Results of a priori sensitivity analyses were provided in order to give insight into the heterogeneity of results.

Conclusions

This study observed that PG is associated with a lower risk for PF compared with PJ. This benefit appeared to be greater in high-risk patients. Surgeons should consider reconstructing the pancreatic remnant following PD with PG, particularly in patients at high risk for PF.

Acknowledgments

The authors would like to thank Dr P. J. Devereaux, Department of Epidemiology and Biostatistics, McMaster University, Hamilton, Ontario, Canada, for sharing results on the Fragility Index measure in order that it could be included in this analysis. They would also like to acknowledge Professor Baki Topal, University Hospitals KU Leuven, Abdominal Surgery, Leuven, Belgium and Dr J. P. Duffas, Digestive Surgery Unit, CHU Rangueil, Toulouse, France, for their responses and for the sharing of data.

Conflicts of interest

None declared.

References

  1. Pellegrini CA, Heck CF, Raper S, Way LW. An analysis of the reduced morbidity and mortality rates after pancreaticoduodenectomy. Arch Surg. 1989;124:778–781. doi: 10.1001/archsurg.1989.01410070028006. [DOI] [PubMed] [Google Scholar]
  2. Crist DW, Sitzmann JV, Cameron JL. Improved hospital morbidity, mortality, and survival after the Whipple procedure. Ann Surg. 1987;206:358–365. doi: 10.1097/00000658-198709000-00014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. DeOliveira ML, Winter JM, Schafer M, Cunningham SC, Cameron JL, Yeo CJ. Assessment of complications after pancreatic surgery: a novel grading system applied to 633 patients undergoing pancreaticoduodenectomy. Ann Surg. 2006;244:931–937. doi: 10.1097/01.sla.0000246856.03918.9a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Berger AC, Howard TJ, Kennedy EP, Sauter PK, Bower-Cherry M, Dutkevitch S. Does type of pancreaticojejunostomy after pancreaticoduodenectomy decrease rate of pancreatic fistula? A randomized, prospective, dual-institution trial. J Am Coll Surg. 2009;208:738–747. doi: 10.1016/j.jamcollsurg.2008.12.031. [DOI] [PubMed] [Google Scholar]
  5. Winter JM, Cameron JL, Campbell KA, Chang DC, Riall TS, Schulick RD. Does pancreatic duct stenting decrease the rate of pancreatic fistula following pancreaticoduodenectomy? Results of a prospective randomized trial. J Gastrointest Surg. 2006;10:1280–1290. doi: 10.1016/j.gassur.2006.07.020. [DOI] [PubMed] [Google Scholar]
  6. Lowy AM, Lee JE, Pisters PW, Davidson BS, Fenoglio CJ, Stanford P. Prospective, randomized trial of octreotide to prevent pancreatic fistula after pancreaticoduodenectomy for malignant disease. Ann Surg. 1997;226:632–641. doi: 10.1097/00000658-199711000-00008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Lillemoe KD, Cameron JL, Kim MP, Campbell KA, Sauter PK, Coleman JA. Does fibrin glue sealant decrease the rate of pancreatic fistula after pancreaticoduodenectomy? Results of a prospective randomized trial. J Gastrointest Surg. 2004;8:766–772. doi: 10.1016/j.gassur.2004.06.011. [DOI] [PubMed] [Google Scholar]
  8. Whipple AO. The rationale of radical surgery for cancer of the pancreas and ampullary region. Cancer Pancreas Ampulla. 1941;114:1–4. doi: 10.1097/00000658-194111440-00008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Topal B, Fieuws S, Aerts R, Weerts J, Feryn T, Roeyen G. Pancreaticojejunostomy versus pancreaticogastrostomy reconstruction after pancreaticoduodenectomy for pancreatic or periampullary tumours: a multicentre randomized trial. Lancet Oncol. 2013;14:655–662. doi: 10.1016/S1470-2045(13)70126-8. [DOI] [PubMed] [Google Scholar]
  10. McKay A, Mackenzie S, Sutherland FR, Bathe OF, Doig C, Dort J. Meta-analysis of pancreaticojejunostomy versus pancreaticogastrostomy reconstruction after pancreaticoduodenectomy. Br J Surg. 2006;93:929–936. doi: 10.1002/bjs.5407. [DOI] [PubMed] [Google Scholar]
  11. Yeo CJ, Cameron JL, Maher MM, Sauter PK, Zahurak MS, Talamini MD. A prospective randomized trial of pancreaticogastrostomy versus pancreaticojejunostomy after pancreaticoduodenectomy. Ann Surg. 1995;222:580–592. doi: 10.1097/00000658-199510000-00014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Wente MN, Shrikhande SV, Müller MW, Diener MK, Seiler CM, Friess H. Pancreaticojejunostomy versus pancreaticogastrostomy: systematic review and meta-analysis. Am J Surg. 2007;193:171–183. doi: 10.1016/j.amjsurg.2006.10.010. [DOI] [PubMed] [Google Scholar]
  13. He T, Zhao Y, Chen Q, Wang X, Lin H, Han W. Pancreaticojejunostomy versus pancreaticogastrostomy after pancreaticoduodenectomy: a systematic review and meta-analysis. Dig Surg. 2013;30:56–69. doi: 10.1159/000350901. [DOI] [PubMed] [Google Scholar]
  14. Shen Y, Jin W. Reconstruction by pancreaticogastrostomy versus pancreaticojejunostomy following pancreaticoduodenectomy: a meta-analysis of randomized controlled trials. Gastroenterol Res Pract. 2012;2012:1–7. doi: 10.1155/2012/627095. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Figueras J, Sabater L, Planellas P, Munoz-Forner E, Lopez-Ben S, Falgueraz L. Randomized clinical trial of pancreaticogastrostomy versus pancreaticojejunostomy on the rate and severity of pancreatic fistula after pancreaticoduodenectomy. Br J Surg. 2013;100:1507–1605. doi: 10.1002/bjs.9252. [DOI] [PubMed] [Google Scholar]
  16. Hallet J, Deobald R, Zih F, Karanicolas PJ, Coburn NG. The impact of pancreatogastrostomy compared to pancreatojejunostomy on postoperative pancreatic fistula following pancreatoduodenectomy: systematic review and meta-analysis. PROSPERO International Prospective Register of Systematic Reviews. Available at http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42013005288 (last accessed 1 April 2014)
  17. Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C, Izbicki J. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery. 2005;138:8–13. doi: 10.1016/j.surg.2005.05.001. [DOI] [PubMed] [Google Scholar]
  18. Higgins JPT, Green S, editors. Cochrane Handbook for Systematic Reviews of Interventions. Chichester: Wiley-Blackwell; 2008. [Google Scholar]
  19. Dindo D, Demartines N, Clavien P-A. Classification of surgical complications. Ann Surg. 2004;240:205–213. doi: 10.1097/01.sla.0000133083.54934.ae. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Koch M, Garden OJ, Padburry R, Rahbari N, Adam R, Capusotti L. Bile leakage after hepatobiliary and pancreatic surgery: a definition and grading of severity by the International Study Group of Liver Surgery. Surgery. 2011;149:680–688. doi: 10.1016/j.surg.2010.12.002. [DOI] [PubMed] [Google Scholar]
  21. Wente MN, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS) Surgery. 2007;142:761–768. doi: 10.1016/j.surg.2007.05.005. [DOI] [PubMed] [Google Scholar]
  22. Balk EM, Bonis PAL, Moskowitz H, Schmid CH, Ioannidis JPA, Wang C. Correlation of quality measures with estimates of treatment effect in meta-analyses of randomized controlled trials. JAMA. 2002;287:2973–2982. doi: 10.1001/jama.287.22.2973. [DOI] [PubMed] [Google Scholar]
  23. Jüni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials for meta-analysis. JAMA. 1999;282:1054–1060. doi: 10.1001/jama.282.11.1054. [DOI] [PubMed] [Google Scholar]
  24. Jüni P, Altman DG, Egger M. Systematic reviews in health care: assessing the quality of controlled clinical trials. BMJ. 2001;323:42–46. doi: 10.1136/bmj.323.7303.42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Bian Z-X, Shang H-C. CONSORT 2010 statement: updated guidelines for reporting parallel group randomized trials. Ann Intern Med. 2011;154:290–291. doi: 10.7326/0003-4819-154-4-201102150-00016. [DOI] [PubMed] [Google Scholar]
  26. Atkins D, Best D, Briss PA, Eccles M, Falck-Ytter Y, Flottorp S. Grading quality of evidence and strength of recommendations. BMJ. 2004;328:1490–1494. doi: 10.1136/bmj.328.7454.1490. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Walsh M, Srinathan S, McAuley DF, Mrkobrada M, Levine O, Ribic C. The statistical significance of randomized controlled trial results: a case for a Fragility Index. J Clin Endocrinol. 2014;67:622–628. doi: 10.1016/j.jclinepi.2013.10.019. [DOI] [PubMed] [Google Scholar]
  28. Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol. 2005;5:13. doi: 10.1186/1471-2288-5-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Higgins JPT, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21:1539–1558. doi: 10.1002/sim.1186. [DOI] [PubMed] [Google Scholar]
  30. Moher D, Liberati A, Tetzlaff J, Altman DG PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;15:264–269. doi: 10.7326/0003-4819-151-4-200908180-00135. [DOI] [PubMed] [Google Scholar]
  31. Fernández-Cruz L, Cosa R, Blanco L, López-Boado MA, Astudillo E. Pancreatogastrostomy with gastric partition after pylorus-preserving pancreatoduodenectomy versus conventional pancreatojejunostomy. Ann Surg. 2008;248:930–938. doi: 10.1097/SLA.0b013e31818fefc7. [DOI] [PubMed] [Google Scholar]
  32. Wellner UF, Brett S, Bruckner T, Limprecht R, Rossion I, Seiler C. Pancreatogastrostomy versus pancreatojejunostomy for RECOnstruction after partial PANCreatoduodenectomy (RECOPANC): study protocol of a randomized controlled trial UTN U1111-1117-9588. Trials. 2012;13:45. doi: 10.1186/1745-6215-13-45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Wellner UF, Kayser G, Lapshyn H, Sick O, Makowiec F, Höppner J. A simple scoring system based on clinical factors related to pancreatic texture predicts postoperative pancreatic fistula preoperatively. HPB. 2010;12:696–702. doi: 10.1111/j.1477-2574.2010.00239.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Kim SW, Youk EG, Park YH. Comparison of pancreatogastrostomy and pancreatojejunostomy after pancreatoduodenectomy performed by one surgeon. World J Surg. 1997;21:640–643. doi: 10.1007/s002689900286. [DOI] [PubMed] [Google Scholar]
  35. Arnaud JP, Tuech JJ, Cervi C, Bergamaschi R. Pancreaticogastrostomy compared with pancreaticojejunostomy after pancreaticoduodenectomy. Eur J Surg. 1999;165:357–362. doi: 10.1080/110241599750006901. [DOI] [PubMed] [Google Scholar]
  36. Takano S, Ito Y, Watanabe Y, Yokoyama T, Kubota N, Iwai S. Pancreaticojejunostomy versus pancreaticogastrostomy in reconstruction following pancreaticoduodenectomy. Br J Surg. 2000;87:423–427. doi: 10.1046/j.1365-2168.2000.01395.x. [DOI] [PubMed] [Google Scholar]
  37. Oussoultzoglou E, Bachellier P, Bigourdan J-M, Weber J-C, Nakano H, Jaeck D. Pancreaticogastrostomy decreased relaparotomy caused by pancreatic fistula after pancreaticoduodenectomy compared with pancreaticojejunostomy. Arch Surg. 2004;139:327–335. doi: 10.1001/archsurg.139.3.327. [DOI] [PubMed] [Google Scholar]
  38. Miyagawa S, Makuuchi M, Lygidakis NJ, Noguchi T, Nishimaki K, Hashikura Y. A retrospective comparative study of reconstructive methods following pancreaticoduodenectomy – pancreaticojejunostomy vs. pancreaticogastrostomy. Hepatogastroenterology. 1992;39:381–384. [PubMed] [Google Scholar]
  39. Matsumoto J, Traverso LW. Exocrine function following the Whipple operation as assessed by stool elastase. J Gastrointest Surg. 2006;10:1225–1229. doi: 10.1016/j.gassur.2006.08.001. [DOI] [PubMed] [Google Scholar]
  40. Balzano G, Zerbi A, Braga M, Rocchetti S, Beneduce AA, Di Carlo V. Fast-track recovery programme after pancreatico-duodenectomy reduces delayed gastric emptying. Br J Surg. 2008;95:1387–1393. doi: 10.1002/bjs.6324. [DOI] [PubMed] [Google Scholar]
  41. Kennedy EP, Rosato EL, Sauter PK, Rosenberg LM, Doria C, Marino IR. Initiation of a critical pathway for pancreaticoduodenectomy at an academic institution – the first step in multidisciplinary team building. J Am Coll Surg. 2007;204:917–923. doi: 10.1016/j.jamcollsurg.2007.01.057. discussion 923–924. [DOI] [PubMed] [Google Scholar]
  42. Callery MP, Pratt WB, Kent TS, Chaikof EL, Vollmer CM. A prospectively validated clinical risk score accurately predicts pancreatic fistula after pancreatoduodenectomy. J Am Coll Surg. 2013;216:1–14. doi: 10.1016/j.jamcollsurg.2012.09.002. [DOI] [PubMed] [Google Scholar]
  43. Ma JP, Peng L, Qin T, Lin JW, Chen C, Cai S. Meta-analysis of pancreaticoduodenectomy prospective controlled trials: pancreaticogastrostomy versus pancreaticojejunostomy reconstruction. Chin Med J. 2012;125:3891–3897. [PubMed] [Google Scholar]
  44. Adie S, Harris IA, Naylor JM, Mittal R. CONSORT compliance in surgical randomized trials: are we there yet? A systematic review. Ann Surg. 2013;258:872–878. doi: 10.1097/SLA.0b013e31829664b9. [DOI] [PubMed] [Google Scholar]
  45. Farrokhyar F, Karanicolas PJ, Thoma A, Simunovic M, Bhandari M, Devereaux PJ. Randomized controlled trials of surgical interventions. Ann Surg. 2010;251:409–416. doi: 10.1097/SLA.0b013e3181cf863d. [DOI] [PubMed] [Google Scholar]

Articles from HPB : The Official Journal of the International Hepato Pancreato Biliary Association are provided here courtesy of Elsevier

RESOURCES