If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Division of Abdominal Surgery, Department of Surgery, Faculty of Medicine, University of Geneva, Geneva, SwitzerlandHepato‐Pancreato‐Biliary Centre, Geneva University Hospitals, Geneva, Switzerland
Division of Abdominal Surgery, Department of Surgery, Faculty of Medicine, University of Geneva, Geneva, SwitzerlandHepato‐Pancreato‐Biliary Centre, Geneva University Hospitals, Geneva, Switzerland
Division of Clinical Epidemiology, Department of Health and Community Medicine, Faculty of Medicine, University of Geneva, Geneva, SwitzerlandCentre de Recherche Clinique, Geneva University Hospitals, Geneva, Switzerland
Division of Abdominal Surgery, Department of Surgery, Faculty of Medicine, University of Geneva, Geneva, SwitzerlandHepato‐Pancreato‐Biliary Centre, Geneva University Hospitals, Geneva, Switzerland
Division of Abdominal Surgery, Department of Surgery, Faculty of Medicine, University of Geneva, Geneva, SwitzerlandHepato‐Pancreato‐Biliary Centre, Geneva University Hospitals, Geneva, Switzerland
Division of Abdominal Surgery, Department of Surgery, Faculty of Medicine, University of Geneva, Geneva, SwitzerlandHepato‐Pancreato‐Biliary Centre, Geneva University Hospitals, Geneva, Switzerland
Division of Abdominal Surgery, Department of Surgery, Faculty of Medicine, University of Geneva, Geneva, SwitzerlandHepato‐Pancreato‐Biliary Centre, Geneva University Hospitals, Geneva, Switzerland
Lymph node involvement in pancreatic adenocarcinoma is a key prognostic factor. Therefore, extending the number of lymph node stations excised in pancreatoduodenectomy may be beneficial to patients with pancreatic adenocarcinoma. This systematic review and meta‐analysis examines the outcomes of extended versus standard lymphadenectomy in the published literature.
Methods
A meta‐analysis of randomized controlled trials (RCTs) comparing extended with standard lymphadenectomy in patients undergoing pancreatoduodenectomy for pancreatic adenocarcinoma was performed. Perioperative outcomes were assessed as pooled odds ratios (ORs) and weighted mean differences. Overall survival was analysed for patients with positive and negative lymph nodes. Results were reported according to the PRISMA statement.
Results
Five RCTs were included, accounting for 724 patients. Extended lymphadenectomy was associated with greater operative time [mean difference: 63 min, 95% confidence interval (CI) 29–96; P < 0.001], increased need for blood transfusions (mean difference: 0.20, 95% CI 0.01–0.30; P = 0.030) and greater postoperative morbidity (OR 1.5, 95% CI 1.25–2.00; P = 0.030), as well as with prolonged diarrhoea after circumferential autonomic nerve dissection around major vessels (OR 12.2, 95% CI 5.3–28.5; P < 0.001). Median survival was similar across the groups in the whole cohort, as well as in subgroups of patients with, respectively, positive and negative lymph nodes.
Conclusions
Extended lymphadenectomy has a harmful impact on patients undergoing oncological pancreatoduodenectomy compared with standard lymphadenectomy.
Introduction
Pancreatic adenocarcinoma is the fourth most common cause of cancer‐related death in high‐income countries.
Complete surgical resection in combination with systemic chemotherapy offers the only chance of potential cure for patients diagnosed with pancreatic adenocarcinoma. Even with such an approach, early recurrence and associated short survival are common outcomes in many patients.
Lymph node involvement is a key prognostic factor in pancreatic adenocarcinoma.
recent evidence has shown that subtle but clinically meaningful subgroups of patients can be distinguished based on the number of positive lymph nodes (PLNs), examined lymph nodes (ELNs), and lymph node ratio [LNR (PLNs divided by ELNs)].
To potentially improve the clearance of infiltrated lymph nodes, extended lymphadenectomy when performing oncological pancreatoduodenectomy (PD) has been proposed by some authors.
have failed to demonstrate an impact of extended lymphadenectomy on perioperative morbidity or survival. However, these systematic reviews are difficult to interpret because they aggregate results gathered from both retrospective and non‐randomized studies. Recently, two a priori powered randomized controlled trials (RCTs) of extended lymphadenectomy in PD have been published.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
The current meta‐analysis of RCTs compares extended (or radical) with standard lymphadenectomy in patients undergoing PD for pancreatic adenocarcinoma and looks at perioperative complications and longterm survival.
Materials and methods
Search strategy and study selection
A systematic review and meta‐analysis was performed according to the PRISMA (preferred reporting items for systematic reviews and meta‐analyses) statement.
Search strategy, trial selection, data extraction and analyses were undertaken according to a predefined protocol, which was registered in the PROSPERO database (CRD 42014013491). The Cochrane Central Register of Controlled Trials, MEDLINE and EMBASE were searched using combinations of terms including ‘pancreatoduodenectomy’, ‘pancreatic adenocarcinoma’, and ‘radical’ or ‘extended’ or ‘standard’ ‘lymphadenectomy’ (The exact search strategy can be found in the Appendix S1, online). Only published studies with an English‐language abstract comparing standard versus extended lymphadenectomy were included. Duplicate entries were excluded. Search results were cross‐referenced with previous meta‐analyses and hand‐searches of the reference lists of retrieved articles were also performed to ensure sensitivity in the identification of relevant RCTs. Two authors (LAO and JM) performed the database searches. Articles were initially screened based on title and abstract. Potentially relevant papers were explored after their full texts had been obtained. To be included, RCTs were required to compare the outcomes of extended lymphadenectomy with those of usual care. It was anticipated that the extent of lymph node excision in both the radical and standard procedures might differ depending on the usual practice of surgeons in different regions of the world. Thus inclusion was not restricted to trials comparing excisions of exactly the same combinations of lymph node stations. Instead, the analysis was planned to include studies comparing a strategy of more aggressive lymph node dissection with usual care, and to investigate clinical variability where required. To ensure comparison worthiness, trials were included only if they reported the number of ELNs in both groups. Quasi‐randomized studies, secondary analyses of preliminary results, and trials that did not provide survival estimates beyond the early postoperative period were excluded.
Data collection and study quality assessment
Two reviewers (LAO and JM) extracted data on study characteristics, baseline clinical variables and outcomes of interest. When several publications nested in the same RCT were available, the most recent and complete estimates were extracted. The quality of each study was assessed according to the Jadad score.
This score contains two items each for randomization and double‐blinding, and one item to evaluate the reporting of cohort attrition. The maximum Jadad score is 5; studies scoring ≥3 are considered of superior quality. Allocation concealment and sample size calculation were also evaluated.
Outcomes of interest
The current meta‐analysis was designed to evaluate three types of outcome: (i) outcomes pertaining to the surgical procedure itself, including operative time and surgical bleeding (as assessed by red blood cell units transfused); (ii) postoperative outcomes, such as morbidity [including the occurrence of overall complications, postoperative pancreatic fistula (POPF), postoperative bleeding, early re‐laparotomy, intra‐abdominal abscess, delayed gastric emptying (DGE), lymphoceles, bile leaks, cholangitis, wound infections], length of hospital stay, and diarrhoea beyond the early postoperative period, and (iii) long term survival and recurrence patterns.
Data analysis
Dichotomous outcomes were pooled as odds ratio (OR) using the Mantel–Haenszel model. For continuous outcomes, the inverse variance method was used to calculate the weighted mean difference (WMD). Heterogeneity was assessed using the Q test and quantified using the I2 statistic. Models with random effects were applied when heterogeneity was detected (P‐value for Q test: <0.100). Survival estimates in the ‘Extended’ and ‘Standard’ study arms were extracted from published survival curves for follow‐up from 6 months to 60 months in steps of 6 months using the R package Digitize (R Foundation for Statistical Computing, Vienna, Austria).
For this purpose, survival curves were digitalized with the package ReadImage (R Foundation for Statistical Computing). Survival estimates were combined as previously described.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
). Combination of survival estimates were conducted for all patients and for patients with positive and negative lymph nodes, respectively. Pooled survival estimates were reported with 95% confidence intervals (95% CIs) and plotted as summary survival curves. Hazard ratios (HRs) were estimated from extracted survival probabilities by calculating relative logarithms of survival for intervals of 6 months (until 60 months).
relative logarithms of survival were combined for each time interval in a single model accounting for random effects (restricted maximum likelihood estimator). As recommended,
a variation of the effect over time was tested in meta‐regression analysis using the R package ‘metafor’ (R Foundation for Statistical Computing). Because the intervention effect was approximately constant over time, a pooled HR over the entire follow‐up was determined. Pooled HRs in patients with, respectively, positive and negative lymph nodes were also compared. A leave‐one‐out sensitivity analysis was conducted. Significance was accepted at the two‐sided 5% level. Analyses were performed using RevMan Version 5.2 (Nordic Cochrane Centre, Copenhagen, Denmark). For some analyses, the R packages ‘msurv’ and ‘mvmeta’ (R Foundation for Statistical Computing) were used.
Results
Study selection
Database searches identified 422 studies (Fig. 1), out of which eight published reports of RCTs were retained.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
Standard vs. radical pancreaticoduodenectomy for periampullary adenocarcinoma: a prospective, randomized trial evaluating quality of life in pancreaticoduodenectomy survivors.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, part 2: randomized controlled trial evaluating survival, morbidity, and mortality.
Pancreaticoduodenectomy with or without extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma: comparison of morbidity and mortality and short‐term outcome.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
Standard vs. radical pancreaticoduodenectomy for periampullary adenocarcinoma: a prospective, randomized trial evaluating quality of life in pancreaticoduodenectomy survivors.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, part 2: randomized controlled trial evaluating survival, morbidity, and mortality.
Pancreaticoduodenectomy with or without extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma: comparison of morbidity and mortality and short‐term outcome.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
As expected, the number of ELNs was significantly higher in the extended lymphadenectomy groups in all study cohorts, with a pooled WMD of 14 lymph nodes (95% CI 10–19; P < 0.001). However, there was some clinical variability in the extent of lymph node clearance. Briefly, the main difference referred to the increased clearance of para‐aortic and coeliac axis lymph nodes in the radical procedures, followed by lymph nodes of the hepatoduodenal ligament (Fig. 2; Japanese classification
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
considered distal gastrectomy as part of the radical procedure only. The remaining studies left the decision on whether or not to perform pylorus‐preserving PD to the discretion of the surgeon.
Figure 2Lymph nodes harvested during (a) standard and (b) extended lymphadenectomy. The colour code indicates numbers of studies harvesting the relevant lymph node groups (Japanese classification). (c) Forest plot illustrating the number of resected lymph nodes (random‐effects model). The vertical line shows the null hypothesis; the surface areas of the blue rectangles indicate the weight of individual studies in the pooled analysis and the black diamond depicts the pooled effect size. 95% CI, 95% confidence interval.
(Adapted from the Japanese Gastric Cancer Association,
Compared with those submitted to standard lymphadenectomy, patients allocated to extended lymphadenectomy experienced significantly prolonged operative time (WMD 63 min, 95% CI 29–96; P < 0.001, Q test P < 0.100, I2 = 79%, random‐effects model), and required marginally more transfusions (WMD 0.20 units, 95% CI 0.01–0.30; P = 0.030, Q test P = 0.690, fixed‐effects model). Overall complications were significantly more common in the extended lymphadenectomy group (OR 1.5, 95% CI 1.1–2.0; P = 0.030, Q test P = 0.200, fixed‐effects model). However, occurrences of single morbidities did not differ statistically significantly between groups (all Q tests, P > 0.100, fixed‐effects model), including for POPF (OR 1.6, 95% CI 0.9–2.7; P = 0.100), bile leaks (OR 3.1, 95% CI 1.0–9.9; P = 0.050), lymphoceles (OR 5.2, 95% CI 0.9–30.4; P = 0.070) and wound infections (OR 1.9, 95% CI 1.0–3.5; P = 0.050). There was no evidence of any differences in occurrences of cholangitis (OR 1.3, 95% CI 0.3–4.8; P = 0.730), DGE (OR 1.5, 95% CI 0.6–3.6; P = 0.400), postoperative bleeding (OR 1.3, 95% CI 0.4–3.9; P = 0.620), re‐laparotomy (OR 1.1, 95% CI 0.4–3.0; P = 0.810), or length of hospital stay (WMD 1 day, 95% CI −2 to +4; P = 0.580) (Figs S1 and S2, online). In studies in which circumferential nerve plexus clearance around the SMA was performed for radical resection,
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
patients had markedly higher odds of experiencing prolonged diarrhoea (OR 12.2, 95% CI 5.3−29.5; P < 0.001, Q test P = 0.100). By contrast, pooled results of studies that involved semi‐circumferential nerve dissection showed diarrhoea to be equally common in the extended and standard lymphadenectomy groups (OR 1.2, 95% CI 0.6–2.5; P = 0.630).
Survival analyses
Pooled summary survival curves are shown in Fig. 3(a). Given the between‐study variability (I2 = 51.4% and 50.0%, respectively), subgroup analyses of patients according to lymph node status (positive and negative, respectively) were also carried out (Fig. 3b and c). Briefly, survival estimates were similar across groups, both in the whole population and within subgroups of patients with positive and negative lymph nodes.
Figure 3Summary survival curves with 95% confidence bands (dashed lines) represented by study groups (black lines for extended lymphadenectomy and grey lines for standard lymphadenectomy) for (a) the overall population, (b) lymph node‐positive patients, and (c) lymph node‐negative patients
To determine whether extended lymphadenectomy impacted on survival, estimated HRs for patients with positive and negative lymph node in each individual study were assessed (Fig. 4). Pooled HRs were not significant for any time interval for patients with either positive or negative lymph nodes and there was no evidence that the HR varied over time (patients with positive and negative lymph node meta‐regression P = 0.184 and P = 0.904, respectively). Therefore, the results were pooled over the entire follow‐up, yielding pooled HRs (extended versus standard lymphadenectomy) of 0.98 (95% CI 0.66–1.46; P = 0.913) and 1.06 (95% CI 0.69–1.64; P = 0.793) for patients with positive and negative lymph nodes. As another approach, the HRs of patients with positive and negative lymph nodes were compared in a meta‐regression with lymph node positivity as a predictor. Again, no statistically significant difference (P = 0.860) was observed, suggesting that extended lymphadenectomy is ineffective in both populations. Finally, when data for both patients with positive and those with negative lymph nodes were combined, the overall pooled HR (extended versus standard lymphadenectomy) was 1.01 (95% CI 0.77–1.34; P = 0.923). In the leave‐one‐out sensitivity analysis, this HR ranged from 0.93 (95% CI 0.72–1.21; P = 0.595) when the study by Nimura et al.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
was removed. The findings were robust with regard to all of the studies.
Figure 4Pooled estimates of hazard ratios (HRs) for each interval of time (black squares) and for the overall follow‐up (grey horizontal line) in (a) lymph node‐positive and (b) lymph node‐negative patients. Vertical black lines represent 95% confidence intervals around the pooled HRs in each interval of time and the grey dashed horizontal line represents the 95% interval around the HR pooled over the entire follow‐up. Circles represent estimated HRs extracted from studies
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
provided detailed tables of recurrence patterns; there was no pooled difference in site of recurrence (local, peritoneal, lymphatic, hepatic) in the extended versus standard lymphadenectomy groups (data not shown).
Discussion
This meta‐analysis reveals that extended lymphadenectomy in PD is associated with increases in operative time, requirements for blood transfusions and incidence of overall complications. With reference to the longterm prognosis, there is no evidence of a benefit of extended lymphadenectomy. Thus, the available evidence suggests that extended lymphadenectomy in PD is a harmful intervention. These findings deserve further comment.
Although intuitively meaningful, extended lymphadenectomy entails increased surgical manipulation, and local cancer cell dissemination may be an unanticipated harmful event when embarking on a radical retroperitoneal dissection. Although in the present analysis, group allocation (extended versus standard lymphadenectomy) was not associated with site of recurrence, only two studies reported recurrence patterns (with an increased rate of peritoneal recurrence after extended lymphadenectomy in one study
Pancreatectomy using the no‐touch isolation technique followed by extensive intraoperative peritoneal lavage to prevent cancer cell dissemination: a pilot study.
By hanging up the pancreatic head using a tape placed along the posterior aspect of the Gerota fascia, ‘no‐touch’ pancreatectomy allows the surgeon to resect the duodenopancreatic bloc without grasping and squeezing the tumour, thus theoretically minimizing cancer cell shedding. A pilot retrospective analysis reported encouraging results
Pancreatectomy using the no‐touch isolation technique followed by extensive intraoperative peritoneal lavage to prevent cancer cell dissemination: a pilot study.
and a recent small RCT (n = 12) reported by Gall et al. demonstrated an increase in the number of circulating cancer cells in the portal blood in 83% of patients following standard resection versus 0% following a no‐touch approach (P = 0.003).
Therefore, although such an approach is conceptually appealing, its clinical effectiveness remains to be formally determined.
Given that most of the trials reported a priori sample size calculations based on longterm survival, that the present survival meta‐analysis pools data from all five studies, and that the estimated HRs were not only non‐significant but also very close to null (extended versus standard lymphadenectomy: HR 1.01, 95% CI 0.77–1.34; P = 0.923), it appears that the lack of association between the extent of lymphadenectomy and survival truly reflects a lack of efficacy of the intervention, rather than residual type II error. However, these observations should be interpreted with some caution because none of the trials included used a non‐inferiority design. Interestingly, the current findings reproduce the evidence surrounding the role of radical lymphadenectomy in gastric cancer surgery. In a recent meta‐analysis, extended lymphadenectomy for proven gastric adenocarcinoma was associated with significantly higher rates of anastomotic leak, overall complications and reoperations, and with 5‐year survival similar to that subsequent to the standard approach.
Diarrhoea, as assessed up to 4 months after surgery, was markedly more common (OR 12.24, 95% CI 5.26–28.47) in patients in whom nerve tissue surrounding the SMA was cleared circumferentially. By contrast, in studies in which only the right lateral aspect of the nerve plexus was dissected, no impact on intestinal transit was observed. Of note, in the trial by Jang et al.,
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
two patients (one of whom subsequently died) in the extended lymphadenectomy group developed SMA pseudoaneurysm after nerve dissection. Thus, nerve plexus dissection not only alters intestinal transit, but also seems to foster the occurrence of vascular complications, which are a common cause of severe morbidity after pancreatectomy.
The strengths of this meta‐analysis include, firstly, its power, which allowed the uncovering of increased postoperative morbidity in the extended lymphadenectomy group, and, secondly, its ability to reconstitute pooled survival curves, as well as to estimate aggregate HRs, which did not indicate that overall survival was affected by the type of lymphadenectomy. However, there are some limitations to the present study. Firstly, given the large number of lymph node stations populating the upper abdominal cavity, some clinical variability is unavoidable when data from trials completed on various continents are pooled. In this respect, clinical variability has been thoroughly inspected (Fig. 2) and it appears that the by far most common differences between standard and extended lymphadenectomy concern para‐aortic and coeliac lymph nodes, as well as nerve plexus dissection around the SMA and coeliac axis. Secondly, although they are informative, subgroup analyses of outcomes in patients with, respectively, positive and negative lymph node patients should be interpreted with caution because randomization allows a random distribution of confounders in the whole study population, but does not guarantee the comparability of specific subgroups (e.g. patients with, respectively, positive and negative lymph nodes).
In conclusion, this meta‐analysis does not support any impact of extended lymphadenectomy on longterm oncological outcomes after PD, but rather demonstrates that extended lymphadenectomy jeopardizes the postoperative course of patients undergoing this procedure. Based on these results, extended lymphadenectomy should not be applied routinely in patients with pancreas head adenocarcinoma requiring PD. However, in an era of patient‐centred medicine, some components of extended PD (e.g. nerve dissection) may still be regarded as beneficial in some patients in whom the tumour is located near the SMA in order to achieve negative‐margin resection.
Conflict of interest
None declared.
Supporting information
Additional Supporting Information may be found in the online version of this article:
Figure S1. Results of pooled analyses of perioperative outcomes: (a) operative time; (b) units of red blood cells transfused; (c) overall complications; (d) postoperative pancreatic fistula (POPF); (e) bile leaks, and (f) lymphoceles. Point estimates are shown with 95% confidence intervals. The vertical line shows the null hypothesis. The surface areas of the blue rectangles indicate the weight of individual studies in the pooled analysis.
Figure S2. Results of pooled analyses of perioperative outcomes: (a) wound complications; (b) postoperative intra‐abdominal bleeding; (c) re‐laparotomy; (d) delayed gastric emptying; (e) cholangitis, and (f) length of hospital stay. Point estimates are shown with 95% confidence intervals. The vertical line shows the null hypothesis. The surface areas of the blue rectangles indicate the weight of individual studies in the pooled analysis.
Appendix S1. Search strategy run in MEDLINE.
Acknowledgements
This study was funded by the Artères Foundation. LAO was supported by the Ligue Genevoise contre le Cancer, and the Dr Henri Dubois‐Ferrière/Dinu Lipatti Foundation. CT and LAO were supported by the Swiss National Science Foundation (grants 323530‐151477 and 3232230‐126233).
A prospective randomized controlled study comparing outcomes of standard resection and extended resection, including dissection of the nerve plexus and various lymph nodes, in patients with pancreatic head cancer.
A prospective randomized trial comparing standard pancreatoduodenectomy with pancreatoduodenectomy with extended lymphadenectomy in resectable pancreatic head adenocarcinoma.
Standard versus extended lymphadenectomy associated with pancreatoduodenectomy in the surgical treatment of adenocarcinoma of the head of the pancreas: a multicenter, prospective, randomized study. Lymphadenectomy Study Group.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma – part 3: update on 5‐year survival.
Standard vs. radical pancreaticoduodenectomy for periampullary adenocarcinoma: a prospective, randomized trial evaluating quality of life in pancreaticoduodenectomy survivors.
Pancreaticoduodenectomy with or without distal gastrectomy and extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma, part 2: randomized controlled trial evaluating survival, morbidity, and mortality.
Pancreaticoduodenectomy with or without extended retroperitoneal lymphadenectomy for periampullary adenocarcinoma: comparison of morbidity and mortality and short‐term outcome.
Pancreatectomy using the no‐touch isolation technique followed by extensive intraoperative peritoneal lavage to prevent cancer cell dissemination: a pilot study.
31234-X&id=gr1.jpg){kind=link}
31234-X&id=gr2.jpg){kind=link}
31234-X&id=gr3.jpg){kind=link}
31234-X&id=gr4.jpg){kind=link}