- Email: [email protected]
A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy
HPB
https://doi.org/10.1016/j.hpb.2019.06.002
ORIGINAL ARTICLE
A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy Jie Cai1,*, Gang Yang2,*, Yun Tao1, Yong Han1, Likai Lin1 & Xinghuan Wang1 1
Hospital Management Institute, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, 430071, China, and 2Department of General Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100730, China
Abstract Background: An appropriate nutritional support is an important consideration for patients undergoing pancreaticoduodenectomy (PD). Recently, early enteral nutrition (EEN) has been considered to be more effective than total parenteral nutrition (TPN) for the early recovery of patients after many digestive tract surgeries. However, there is little evidence to support EEN in patients undergoing PD. Methods: A systematic literature review was performed to identify relevant studies before December 2018. Statistical analysis was carried out using Review Manager 5.3. Results: Nine studies with 1258 patients were included in the meta-analysis. Six studies compared EEN and TPN and three compared two strategies combined vs. a single strategy. The length of hospital stay (LOS) in the EEN group was significantly shorter than that in the TPN group (P < 0.001). There was no difference in the risk of postoperative complications, infections, and mortality between the EEN and TPN groups. In the comparison of two combined strategies vs. one, no significant difference was seen in overall postoperative complications, LOS, or mortality. Conclusion: Compared with TPN, EEN is a safe strategy and can substantially shorten the LOS of patients. Received 7 January 2019; accepted 16 June 2019
Correspondence Xinghuan Wang; Likai Lin.
E-mails: [email protected] (L. Lin), [email protected]
(X. Wang)
Introduction Pancreaticoduodenectomy (PD), a standard treatment for periampullary cancer, is a high-risk surgical intervention. Although mortality has declined to less than 5%, PD is still accompanied by a high postoperative incidence (up to 60%) of various complications, including postoperative pancreatic fistula (POPF), delayed gastric emptying (DGE).1 Patients with periampullary cancer are often malnourished, which can affect the healing process, intestinal barrier function and can increase postoperative complications. Therefore, perioperative nutritional supplementation, including early enteral nutrition (EEN) and total parenteral nutrition (TPN), are important for reducing *
These authors contributed equally to this study.
HPB xxxx, xxx, xxx
complications and promoting recovery after PD surgery, as recommended by guidelines of the European Society for Clinical Nutrition and Metabolism.2,3 Numerous studies have shown EEN to have a positive influence on reducing the incidence of postoperative complications, following digestive tract surgery, and EEN is more physiologically and economically efficient than TPN.4,5 In particular, with the emergence of enhanced recovery after surgery (ERAS), increasingly more surgeons advocate EEN after surgery, including early oral food intake after digestive tract surgery.6–8 However, owing to the complexity and high risk of PD surgery, the best route of nutritional support after PD remains controversial. Some randomized controlled trials (RCT) have revealed that EEN is superior to TPN for patients with pancreatic cancer who undergo PD.9,10 However, other studies indicate that EEN does not
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
2
provide as much benefit as expected in terms of POPF, infectious complications, and length of stay (LOS) after PD. In addition, some patients cannot tolerate enteral nutrition (EN) owing to diarrhea, gastric spasm, and other symptoms.11,12 What’s more, some studies have revealed that patients receiving EEN have a higher complication rate, especially those with poorer nutritional status, in comparison with patients receiving TPN.13,14 These contradictory results can affect clinicians’ decisionmaking regarding nutritional strategies after PD. To clarify the benefits of EEN and TPN in patients undergoing PD, the present meta-analysis was performed to determine the best method of nutritional support after PD.
Methods The study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.15 Study selection The search for relevant studies was performed in three databases, PubMed, Embase, and the Cochrane Library, using the key words “pancreaticoduodenectomy”, “Whipple procedure” or “Whipple operation”, “enteral nutrition” or “early enteral nutrition”, and “total parenteral nutrition” or “parenteral nutrition”. The term “OR” was used as the set operator to combine different sets of results. When multiple articles for a similar study were identified, only the most complete and recently published ones were considered. References from relevant reviews and abstracts were also consulted, to identify additional studies. The last date of study selection was on 24 December 2018. The inclusion criteria were as follows: 1) RCT or Non-RCT (NRCT) with parallel-controlled design; 2) patients underwent PD; 3) the trial groups consisted of at least two of EN group, TPN group, or EN combined with PN group; 4) EN was started prior to 48 h after surgery; 5) main outcomes such as postoperative complications, postoperative hemorrhage, postoperative infections, and LOS were described; 6) studies were published in English and the relevant data were available. The following types of study were excluded: 1) duplicate publication; 2) noncomparative studies; 3) animal trials, non-related studies, review articles, and case reports. Data extraction and assessment of risk of bias Included studies were collected, coded, and evaluated independently by two investigators (CJ and YG), according to the PRISMA guidelines; differing opinions between reviewers were resolved by discussion. The following information was extracted from each study: 1) general information, such as name of the first author, year of publication, sample size, design, method of EN, and when EN was started; 2) outcomes such as morbidity of DGE, postoperative hemorrhage, POPF, biliary
HPB xxxx, xxx, xxx
fistula, intra-abdominal infection, wound infection, and lung infection, as well as overall morbidity of complications, mortality, and LOS. RCTs were evaluated using the Jadad composite scale.16 The top score on the Jadad scale is 5, and studies with scores of 3 or more were considered to be of high quality. NRCTs were assessed using the Newcastle–Ottawa quality assessment scale,17 in which the top score is 9; studies with scores of 6 or greater are considered high quality. Data analysis The data were analyzed following classical meta-analysis method using Review Manager (RevMan) Version 5.3. (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). Risk ratios (RR) with 95% confidence intervals (CIs) were obtained after analyzing dichotomous data, and the weighted mean difference with 95% CI was obtained by synthesizing continuous data. A statistical test for heterogeneity was performed using the I2 statistic. When there was significant heterogeneity (I2 50%) between studies, a random-effects model was used; otherwise, a fixed-effects model was used. P < 0.05 was considered to be statistically significant.
Results Literature search Nine studies including 1258 patients were included in the metaanalysis (Fig. 1).9,10,14,18–23 Detailed information of these 9 studies is provided in Table 1. Six studies including 727 patients were grouped as studies comparing EEN (n = 397) versus TPN (n = 330); three studies including 531 patients were grouped as those comparing EEN and PN (n = 329) versus EEN or PN (n = 202). Five studies were RCTs, and the remaining four were all retrospective cohort studies. Meta-analysis results of EEN versus TPN after PD Ten items (overall postoperative complications, DGE, postoperative hemorrhage, POPF, mortality, LOS, biliary fistula, intra-abdominal infection, wound infection and lung infection) were analyzed in a comparison of outcomes between patients who received EEN and those who received TPN. Sub-groups in the original research, such as different methods and types of EEN, were merged using Revman 5.3. After analysis, the mean difference in LOS between the two groups was significant (P < 0.001). The final mean difference was −1.46 (95% CI −2.04, −0.89) days, which meant that LOS in the EEN group was shorter than that of the TPN group (Fig. 2). However, no significant difference was shown in other complications such as DGE (Fig. 3), POPF (Fig. 4) and so on between two groups (P > 0.05). The detailed data are shown in Table 2.
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
3
Discussion This study finally comprised nine studies, including five RCTs and four NRCTs. In view of the actual clinical situation, all relevant trials could not satisfy the double-blinding principle for EEN or TPN therapy. Although not all of the included studies were RCTs, eight out of nine studies included in this metaanalysis were assessed to be high quality, which could ensure the reliability of the meta-analysis. In recent years, owing to the promotion of ERAS, one of its important components, EEN, has been increasingly adopted in digestive tract surgery.3,24,25 However, owing to complexity and difficulty, some surgeons remain very cautious about the choice of postoperative nutritional strategies for patients following PD, especially as some studies have indicated that EEN could increase the rate of overall postoperative complications and POPF after PD.14 The present meta-analysis showed that EEN is as tolerable and safe as TPN for patients after PD, with no increase in DGE (RR 1.88; 95% CI 0.55, 6.47; P = 0.32), POPF (RR 1.33; 95% CI 0.97, 1.84; P = 0.08), intra-abdominal infection (RR 0.62; 95% CI 0.26, 1.49; P = 0.28), and mortality (RR 0.43; 95% CI 0.11, 1.62; P = 0.21). However, EEN can significantly reduce LOS (RR −1.46; 95% CI −2.04, −0.89; P < 0.001). Although hospitalization cost was not included in this study, the authors can reasonably speculate that costs for the EEN group were lower, based on the fact that EEN is less expensive and LOS was shorter in the EEN group. Prior to this study, a meta-analysis of four RCTs by Shen et al. examined the effect of EEN in patients after PD.26 Those authors reported that EEN appeared safe and was tolerated by patients after PD, but EEN did not show advantages with respect to
Figure 1 Study selection flow chart
Meta-analysis results of EEN and PN versus EEN or TPN after PD In this analysis, five items (POPF, wound infection, lung infection, mortality and LOS) were compared for outcomes between the EEN + PN group and EEN/TPN group after PD. Three studies with 531 patients were included in each item.21–23 There was no significant difference in the comparison of all items, and detailed data are shown in Table 3.
Table 1 Basic characteristics of studies included in the meta-analysis
Author
Year Country
Sample size
Liu et al.
2011 China
28
Design
Way of EN
When EN started (hours after surgery)
Study-quality Score
EEN TPN EEN + PN –
RCT
jejunostomy tube
48h
3
Perinel et al. 2016 France
103 101 –
RCT
nasojejunal tube
48h
3
Gianotti et al.
2000 Italy
144 68
–
RCT
jejunostomy tube
12h
2
Park et al.
2012 Korea
18
20
–
RCT
nasojejunal tube
24h
3
Gerritsen et al.
2012 Netherlands 92
37
–
retrospective cohort
44 nasojejunal tube 48 jejunostomy tube
24h
8
Guilbaud et al.
2017 France
12
74
–
retrospective cohort
jejunostomy tube
24h
9
Zhu et al.
2013 China
–
107 67
retrospective cohort
nasojejunal tube
24h
8
Lu et al.
2016 China
–
87
253
retrospective cohort
nasojejunal tube
24h
8
Nagata et al.
2009 Japan
8
–
9
RCT
jejunostomy tube
24h
3
30
EEN, Early Enteral Nutrition; EN, Enteral Nutrition; PN, Parenteral Nutrition; TPN, Total Parenteral Nutrition; RCT, Randomised Controlled Trial.
HPB xxxx, xxx, xxx
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
4
Figure 2 Length of hospital stay in early enteral nutrition (EEN) vs. total parenteral nutrition (TPN)
Figure 3 Delayed gastric emptying morbidity in EEN vs. TPN
infection and LOS. Although all studies included in that metaanalysis were RCTs, only two compared EEN and TPN groups. Patients in a study by Tien et al. were grouped according to different styles of surgery, with corresponding nutritional strategies.27 The impact of different surgical methods on outcomes might conceal the impact of nutritional strategies on outcomes, thus interfering with the analysis. The result of a literature review published by Buscemi et al.12 was consistent with that of Shen et al..26 Different from their result, LOS was shorter in the EEN group than in the TPN group in the current study, which is of importance to cost reduction and patient recovery. In addition to simple EEN or TPN, some researchers have attempted to combine these two strategies, to alleviate EENrelated complications such as diarrhea, abdominal distention, and abdominal cramps.21–23 However, Lu et al. reported that in
patients after PD, EEN combined with PN showed higher risk for DGE and pulmonary infection than TPN.22 The present metaanalysis indicated that there was no significant difference in the risk of POPF, wound and lung infection, mortality, and LOS for one simple nutrition strategy or a combined approach. However, more studies are needed to confirm this conclusion owing to the lack of high-quality RCTs for this subgroup. The authors hope that comparison of a combination of two nutritional methods with a single approach can provide a reference for researchers in conducting clinical experiments and meta-analyses in this area in the future. The relatively small quantity of included studies is the main limitation of the current study, especially in the analysis between one nutrition strategy group and two strategies combined. For this reason, funnel plots of publishing bias were not created. In
Figure 4 Pancreatic fistula morbidity in EEN vs. TPN
HPB xxxx, xxx, xxx
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
5
Table 2 Meta-analysis result of EEN vs. TPN after PD
Outcome
Studies
Participants
Statistical Method
Effect Estimate
P value
I2
Overall postoperative complications
5
669
RR (M-H, Random)
0.95 [0.60, 1.50]
0.84
78%
Delayed gastric emptying
6
721
RR (M-H, Fixed)
0.99 [0.91, 1.08]
0.80
2%
Postoperative hemorrhage
6
726
RR (M-H, Fixed)
0.86 [0.58, 1.27]
0.45
20%
Pancreatic fistula
6
726
RR (M-H, Fixed)
1.33 [0.97, 1.84]
0.08
46%
Mortality
5
689
RR (M-H, Random)
1.88 [0.55, 6.47]
0.32
55%
Length of hospital stay (day)
5
603
MD (IV, Fixed)
−1.46 [−2.04, −0.89]
<0.001
20%
Biliary fistula
4
485
RR (M-H, Fixed)
0.96 [0.40, 2.33]
0.94
0%
Intra-abdominal infection
3
399
RR (M-H, Fixed)
0.62 [0.26, 1.49]
0.28
0%
Wound infection
3
379
RR (M-H, Fixed)
0.53 [0.29, 1.00]
0.05
26%
Lung infection
3
399
RR (M-H, Fixed)
1.03 [0.47, 2.24]
0.94
33%
EEN: Early Enteral Nutrition; TPN, Total Parenteral Nutrition; PD, Pancreaticoduodenectomy; RR, Risk Ratio.
Table 3 Meta-analysis result of EEN + PN vs. EEN/TPN after PD
Outcome
Studies
Participants
Statistical Method
Effect Estimate
P value
I2
Pancreatic fistula
3
531
RR (M-H, Fixed)
1.42 [0.80, 2.50]
0.23
0%
Wound infection
3
531
RR (M-H, Fixed)
0.83 [0.48, 1.44]
0.50
0%
Lung infection
3
531
RR (M-H, Random)
0.95 [0.20, 4.38]
0.95
75%
Mortality
3
531
RR (M-H, Fixed)
0.95 [0.24, 3.83]
0.94
0%
Length of hospital stay (day)
3
531
MD (IV, Random)
5.64 [−1.80, 13.09]
0.14
98%
the article by Gerritsen et al.,19 the standard deviation of continuous variables was not reported and could not contribute to the meta-analysis, which might affect the outcome. In addition, owing to non-availability of the original data on demographic information, disease, and surgical method, subgroup analysis could not be performed, making it unclear whether EEN is potentially more effective or risky in these subgroups. Data of subgroups in the EEN group were also included, such as immune-enhanced EEN and standard EEN or nasojejunal tube and jejunostomy tube. These factors might influence the effect of EEN; however, few studies reported outcomes, making our further research in this regard impossible. In this research, EEN after PD was compared with TPN after PD, and EEN was found to reduce the time and cost of hospitalization. Therefore, EEN might be a more appropriate choice. After combining two nutrition strategies, no obvious advantages or disadvantages were observed, suggesting that such combination was safe and could be used to supplement a single nutritional strategy, especially for EN-intolerant patients.
Authors’ contributions Conceived and designed the experiments: CJ and YG. Literature search and selection: CJ and YG. Analyzed the data: CJ. Contributed reagents/materials/analysis tools: TY, LL and WX. Wrote the paper: CJ and YG. All authors have read and approved the manuscript.
Funding This study was supported by PUMC Youth Fund and the Fundamental Research Funds for the Central Universities [grant number 2017320027] and the Chinese Academy Medical Science Innovation Fund for Medical Students [grant number 20171002-1-16].
Conflicts of interest None. References 1. Schnelldorfer T, Ware AL, Sarr MG, Smyrk TC, Zhang L, Qin R et al. (2008) Long-term survival after pancreatoduodenectomy for pancreatic
Conclusion
adenocarcinoma: is cure possible? Ann Surg 247:456–462. 2. Weimann A, Braga M, Harsanyi L, Laviano A, Ljungqvist O, Soeters P
Compared with TPN, EEN is a safe strategy and can significantly shorten the LOS in patients after PD. For patients with gastrointestinal tolerance, EEN can be considered a priority; otherwise, PN combined with EEN is also a safe strategy.
HPB xxxx, xxx, xxx
et al. (2006) ESPEN guidelines on enteral nutrition: surgery including organ transplantation. Clin Nutr 25:224–244. 3. Weimann A, Braga M, Carli F, Higashiguchi T, Hübner M, Klek S et al. (2017) ESPEN guideline: clinical nutrition in surgery. Clin Nutr 36:623–650.
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
6
4. Bozzetti F, Braga M, Gianotti L, Gavazzi C, Mariani L. (2001) Postoperative
17. Stang A. (2010) Critical evaluation of the Newcastle-Ottawa scale for the
enteral versus parenteral nutrition in malnourished patients with gastroin-
assessment of the quality of nonrandomized studies in meta-analyses.
testinal cancer: a randomised multicentre trial. Lancet 358:1487–1492.
Eur J Epidemiol 25:603–605.
5. Mazaki T, Ebisawa K. (2008) Enteral versus parenteral nutrition after
18. Park JS, Chung H, Hwang HK, Kim JK, Yoon DS. (2012) Postoperative
gastrointestinal surgery: a systematic review and meta-analysis of
nutritional effects of early enteral feeding compared with total parental
randomized controlled trials in the English literature. J Gastrointest Surg
nutrition in pancreaticoduodectomy patients: a prosepective, randomized study. J Korean Med Sci 27:261.
12:739–755. 6. Group EC. (2015) The impact of enhanced recovery protocol compli-
19. Gerritsen A, Besselink MG, Cieslak KP, Vriens MR, Steenhagen E, van
ance on elective colorectal cancer resection: results from an Interna-
Hillegersberg R et al. (2012) Efficacy and complications of nasojejunal,
tional registry. Ann Surg 261:1153–1159.
jejunostomy and parenteral feeding after pancreaticoduodenectomy.
7. Visioni A, Shah R, Gabriel E, Attwood K, Kukar M, Nurkin S. (2018)
J Gastrointest Surg 16:1144–1151.
Enhanced recovery after surgery for noncolorectal surgery?: a sys-
20. Guilbaud T, Birnbaum DJ, Loubière S, Bonnet J, Chopinet S, Grégoire E
tematic review and meta-analysis of major abdominal surgery. Ann Surg
et al. (2017) Comparison of different feeding regimes after pancreatoduodenectomy – a retrospective cohort analysis. Nutr J 16.
267:57–65. 8. Gerritsen A, Wennink RA, Besselink MG, van Santvoort HC, Tseng DS,
21. Zhu X. (2013) Effect of early enteral combined with parenteral nutrition in
Steenhagen E et al. (2014) Early oral feeding after pancreatoduodenectomy
patients undergoing pancreaticoduodenectomy. World J Gastroenterol
enhances recovery without increasing morbidity. HPB 16:656–664.
19:5889.
9. Liu C, Du Z, Lou C, Wu C, Yuan Q, Wang J et al. (2011) Enteral nutrition is
22. Lu J, Liu C, Du Z, Liu X, Lv Y, Zhang X. (2016) Early enteral nutrition vs
superior to total parenteral nutrition for pancreatic cancer patients who
parenteral nutrition following pancreaticoduodenectomy: experience
underwent pancreaticoduodenectomy. Asia Pac J Clin Nutr 20:154–160.
from a single center. World J Gastroenterol 22:3821.
10. Gianotti L, Braga M, Gentilini O, Balzano G, Zerbi A, Di Carlo V. (2000)
23. Nagata S, Fukuzawa K, Iwashita Y, Kabashima A, Kinoshita T,
Artificial nutrition after pancreaticoduodenectomy. Pancreas 21:344–351.
Wakasugi K et al. (2009) Comparison of enteral nutrition with combined
11. Seike J, Tangoku A, Yuasa Y, Okitsu H, Kawakami Y, Sumitomo M.
enteral and parenteral nutrition in post-pancreaticoduodenectomy pa-
(2011) The effect of nutritional support on the immune function in the acute postoperative period after esophageal cancer surgery: total parenteral nutrition versus enteral nutrition. J Med Investig 58:75–80. 12. Buscemi S, Damiano G, Palumbo V, Spinelli G, Ficarella S, Monte G et al. (2015) Enteral nutrition in pancreaticoduodenectomy: a literature
tients: a pilot study. Nutr J 8:24. 24. Han H, Pan M, Tao Y, Liu R, Huang Z, Piccolo K et al. (2018) Early enteral nutrition is associated with faster post-esophagectomy recovery in Chinese esophageal cancer patients: a retrospective cohort study. Nutr Cancer 70:221–228. 25. Li B, Liu HY, Guo SH, Sun P, Gong FM, Jia BQ. (2015) Impact of early
review. Nutrients 7:3154–3165. 13. Heyland DK, Montalvo M, MacDonald S, Keefe L, Su XY, Drover JW. (2001) Total parenteral nutrition in the surgical patient: a meta-analysis.
postoperative enteral nutrition on clinical outcomes in patients with gastric cancer. Genet Mol Res 14:7136–7141. 26. Shen Y, Jin W. (2013) Early enteral nutrition after pancreatoduodenec-
Can J Surg 44:102–111. 14. Perinel J, Mariette C, Dousset B, Sielezneff I, Gainant A, Mabrut J et al. (2016) Early enteral versus total parenteral nutrition in patients undergoing pancreaticoduodenectomy. Ann Surg 264:731–737. 15. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP et al. (2009) The PRISMA statement for reporting systematic reviews
tomy: a meta-analysis of randomized controlled trials. Langenbeck’s Arch Surg 398:817–823. 27. Tien Y, Yang C, Wu Y, Hu R, Lee P. (2009) Enteral nutrition and biliopancreatic diversion effectively minimize impacts of gastroparesis after pancreaticoduodenectomy. J Gastrointest Surg 13:929–937.
and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 339:b2700. 16. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ et al. (1996) Assessing the quality of reports of random-
Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10. 1016/j.hpb.2019.06.002.
ized clinical trials: is blinding necessary? Control Clin Trials 17:1–12.
HPB xxxx, xxx, xxx
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
https://doi.org/10.1016/j.hpb.2019.06.002
ORIGINAL ARTICLE
A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy Jie Cai1,*, Gang Yang2,*, Yun Tao1, Yong Han1, Likai Lin1 & Xinghuan Wang1 1
Hospital Management Institute, Zhongnan Hospital of Wuhan University, Wuhan, Hubei, 430071, China, and 2Department of General Surgery, Peking Union Medical College Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100730, China
Abstract Background: An appropriate nutritional support is an important consideration for patients undergoing pancreaticoduodenectomy (PD). Recently, early enteral nutrition (EEN) has been considered to be more effective than total parenteral nutrition (TPN) for the early recovery of patients after many digestive tract surgeries. However, there is little evidence to support EEN in patients undergoing PD. Methods: A systematic literature review was performed to identify relevant studies before December 2018. Statistical analysis was carried out using Review Manager 5.3. Results: Nine studies with 1258 patients were included in the meta-analysis. Six studies compared EEN and TPN and three compared two strategies combined vs. a single strategy. The length of hospital stay (LOS) in the EEN group was significantly shorter than that in the TPN group (P < 0.001). There was no difference in the risk of postoperative complications, infections, and mortality between the EEN and TPN groups. In the comparison of two combined strategies vs. one, no significant difference was seen in overall postoperative complications, LOS, or mortality. Conclusion: Compared with TPN, EEN is a safe strategy and can substantially shorten the LOS of patients. Received 7 January 2019; accepted 16 June 2019
Correspondence Xinghuan Wang; Likai Lin.
E-mails: [email protected] (L. Lin), [email protected]
(X. Wang)
Introduction Pancreaticoduodenectomy (PD), a standard treatment for periampullary cancer, is a high-risk surgical intervention. Although mortality has declined to less than 5%, PD is still accompanied by a high postoperative incidence (up to 60%) of various complications, including postoperative pancreatic fistula (POPF), delayed gastric emptying (DGE).1 Patients with periampullary cancer are often malnourished, which can affect the healing process, intestinal barrier function and can increase postoperative complications. Therefore, perioperative nutritional supplementation, including early enteral nutrition (EEN) and total parenteral nutrition (TPN), are important for reducing *
These authors contributed equally to this study.
HPB xxxx, xxx, xxx
complications and promoting recovery after PD surgery, as recommended by guidelines of the European Society for Clinical Nutrition and Metabolism.2,3 Numerous studies have shown EEN to have a positive influence on reducing the incidence of postoperative complications, following digestive tract surgery, and EEN is more physiologically and economically efficient than TPN.4,5 In particular, with the emergence of enhanced recovery after surgery (ERAS), increasingly more surgeons advocate EEN after surgery, including early oral food intake after digestive tract surgery.6–8 However, owing to the complexity and high risk of PD surgery, the best route of nutritional support after PD remains controversial. Some randomized controlled trials (RCT) have revealed that EEN is superior to TPN for patients with pancreatic cancer who undergo PD.9,10 However, other studies indicate that EEN does not
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
2
provide as much benefit as expected in terms of POPF, infectious complications, and length of stay (LOS) after PD. In addition, some patients cannot tolerate enteral nutrition (EN) owing to diarrhea, gastric spasm, and other symptoms.11,12 What’s more, some studies have revealed that patients receiving EEN have a higher complication rate, especially those with poorer nutritional status, in comparison with patients receiving TPN.13,14 These contradictory results can affect clinicians’ decisionmaking regarding nutritional strategies after PD. To clarify the benefits of EEN and TPN in patients undergoing PD, the present meta-analysis was performed to determine the best method of nutritional support after PD.
Methods The study was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.15 Study selection The search for relevant studies was performed in three databases, PubMed, Embase, and the Cochrane Library, using the key words “pancreaticoduodenectomy”, “Whipple procedure” or “Whipple operation”, “enteral nutrition” or “early enteral nutrition”, and “total parenteral nutrition” or “parenteral nutrition”. The term “OR” was used as the set operator to combine different sets of results. When multiple articles for a similar study were identified, only the most complete and recently published ones were considered. References from relevant reviews and abstracts were also consulted, to identify additional studies. The last date of study selection was on 24 December 2018. The inclusion criteria were as follows: 1) RCT or Non-RCT (NRCT) with parallel-controlled design; 2) patients underwent PD; 3) the trial groups consisted of at least two of EN group, TPN group, or EN combined with PN group; 4) EN was started prior to 48 h after surgery; 5) main outcomes such as postoperative complications, postoperative hemorrhage, postoperative infections, and LOS were described; 6) studies were published in English and the relevant data were available. The following types of study were excluded: 1) duplicate publication; 2) noncomparative studies; 3) animal trials, non-related studies, review articles, and case reports. Data extraction and assessment of risk of bias Included studies were collected, coded, and evaluated independently by two investigators (CJ and YG), according to the PRISMA guidelines; differing opinions between reviewers were resolved by discussion. The following information was extracted from each study: 1) general information, such as name of the first author, year of publication, sample size, design, method of EN, and when EN was started; 2) outcomes such as morbidity of DGE, postoperative hemorrhage, POPF, biliary
HPB xxxx, xxx, xxx
fistula, intra-abdominal infection, wound infection, and lung infection, as well as overall morbidity of complications, mortality, and LOS. RCTs were evaluated using the Jadad composite scale.16 The top score on the Jadad scale is 5, and studies with scores of 3 or more were considered to be of high quality. NRCTs were assessed using the Newcastle–Ottawa quality assessment scale,17 in which the top score is 9; studies with scores of 6 or greater are considered high quality. Data analysis The data were analyzed following classical meta-analysis method using Review Manager (RevMan) Version 5.3. (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). Risk ratios (RR) with 95% confidence intervals (CIs) were obtained after analyzing dichotomous data, and the weighted mean difference with 95% CI was obtained by synthesizing continuous data. A statistical test for heterogeneity was performed using the I2 statistic. When there was significant heterogeneity (I2 50%) between studies, a random-effects model was used; otherwise, a fixed-effects model was used. P < 0.05 was considered to be statistically significant.
Results Literature search Nine studies including 1258 patients were included in the metaanalysis (Fig. 1).9,10,14,18–23 Detailed information of these 9 studies is provided in Table 1. Six studies including 727 patients were grouped as studies comparing EEN (n = 397) versus TPN (n = 330); three studies including 531 patients were grouped as those comparing EEN and PN (n = 329) versus EEN or PN (n = 202). Five studies were RCTs, and the remaining four were all retrospective cohort studies. Meta-analysis results of EEN versus TPN after PD Ten items (overall postoperative complications, DGE, postoperative hemorrhage, POPF, mortality, LOS, biliary fistula, intra-abdominal infection, wound infection and lung infection) were analyzed in a comparison of outcomes between patients who received EEN and those who received TPN. Sub-groups in the original research, such as different methods and types of EEN, were merged using Revman 5.3. After analysis, the mean difference in LOS between the two groups was significant (P < 0.001). The final mean difference was −1.46 (95% CI −2.04, −0.89) days, which meant that LOS in the EEN group was shorter than that of the TPN group (Fig. 2). However, no significant difference was shown in other complications such as DGE (Fig. 3), POPF (Fig. 4) and so on between two groups (P > 0.05). The detailed data are shown in Table 2.
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
3
Discussion This study finally comprised nine studies, including five RCTs and four NRCTs. In view of the actual clinical situation, all relevant trials could not satisfy the double-blinding principle for EEN or TPN therapy. Although not all of the included studies were RCTs, eight out of nine studies included in this metaanalysis were assessed to be high quality, which could ensure the reliability of the meta-analysis. In recent years, owing to the promotion of ERAS, one of its important components, EEN, has been increasingly adopted in digestive tract surgery.3,24,25 However, owing to complexity and difficulty, some surgeons remain very cautious about the choice of postoperative nutritional strategies for patients following PD, especially as some studies have indicated that EEN could increase the rate of overall postoperative complications and POPF after PD.14 The present meta-analysis showed that EEN is as tolerable and safe as TPN for patients after PD, with no increase in DGE (RR 1.88; 95% CI 0.55, 6.47; P = 0.32), POPF (RR 1.33; 95% CI 0.97, 1.84; P = 0.08), intra-abdominal infection (RR 0.62; 95% CI 0.26, 1.49; P = 0.28), and mortality (RR 0.43; 95% CI 0.11, 1.62; P = 0.21). However, EEN can significantly reduce LOS (RR −1.46; 95% CI −2.04, −0.89; P < 0.001). Although hospitalization cost was not included in this study, the authors can reasonably speculate that costs for the EEN group were lower, based on the fact that EEN is less expensive and LOS was shorter in the EEN group. Prior to this study, a meta-analysis of four RCTs by Shen et al. examined the effect of EEN in patients after PD.26 Those authors reported that EEN appeared safe and was tolerated by patients after PD, but EEN did not show advantages with respect to
Figure 1 Study selection flow chart
Meta-analysis results of EEN and PN versus EEN or TPN after PD In this analysis, five items (POPF, wound infection, lung infection, mortality and LOS) were compared for outcomes between the EEN + PN group and EEN/TPN group after PD. Three studies with 531 patients were included in each item.21–23 There was no significant difference in the comparison of all items, and detailed data are shown in Table 3.
Table 1 Basic characteristics of studies included in the meta-analysis
Author
Year Country
Sample size
Liu et al.
2011 China
28
Design
Way of EN
When EN started (hours after surgery)
Study-quality Score
EEN TPN EEN + PN –
RCT
jejunostomy tube
48h
3
Perinel et al. 2016 France
103 101 –
RCT
nasojejunal tube
48h
3
Gianotti et al.
2000 Italy
144 68
–
RCT
jejunostomy tube
12h
2
Park et al.
2012 Korea
18
20
–
RCT
nasojejunal tube
24h
3
Gerritsen et al.
2012 Netherlands 92
37
–
retrospective cohort
44 nasojejunal tube 48 jejunostomy tube
24h
8
Guilbaud et al.
2017 France
12
74
–
retrospective cohort
jejunostomy tube
24h
9
Zhu et al.
2013 China
–
107 67
retrospective cohort
nasojejunal tube
24h
8
Lu et al.
2016 China
–
87
253
retrospective cohort
nasojejunal tube
24h
8
Nagata et al.
2009 Japan
8
–
9
RCT
jejunostomy tube
24h
3
30
EEN, Early Enteral Nutrition; EN, Enteral Nutrition; PN, Parenteral Nutrition; TPN, Total Parenteral Nutrition; RCT, Randomised Controlled Trial.
HPB xxxx, xxx, xxx
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
4
Figure 2 Length of hospital stay in early enteral nutrition (EEN) vs. total parenteral nutrition (TPN)
Figure 3 Delayed gastric emptying morbidity in EEN vs. TPN
infection and LOS. Although all studies included in that metaanalysis were RCTs, only two compared EEN and TPN groups. Patients in a study by Tien et al. were grouped according to different styles of surgery, with corresponding nutritional strategies.27 The impact of different surgical methods on outcomes might conceal the impact of nutritional strategies on outcomes, thus interfering with the analysis. The result of a literature review published by Buscemi et al.12 was consistent with that of Shen et al..26 Different from their result, LOS was shorter in the EEN group than in the TPN group in the current study, which is of importance to cost reduction and patient recovery. In addition to simple EEN or TPN, some researchers have attempted to combine these two strategies, to alleviate EENrelated complications such as diarrhea, abdominal distention, and abdominal cramps.21–23 However, Lu et al. reported that in
patients after PD, EEN combined with PN showed higher risk for DGE and pulmonary infection than TPN.22 The present metaanalysis indicated that there was no significant difference in the risk of POPF, wound and lung infection, mortality, and LOS for one simple nutrition strategy or a combined approach. However, more studies are needed to confirm this conclusion owing to the lack of high-quality RCTs for this subgroup. The authors hope that comparison of a combination of two nutritional methods with a single approach can provide a reference for researchers in conducting clinical experiments and meta-analyses in this area in the future. The relatively small quantity of included studies is the main limitation of the current study, especially in the analysis between one nutrition strategy group and two strategies combined. For this reason, funnel plots of publishing bias were not created. In
Figure 4 Pancreatic fistula morbidity in EEN vs. TPN
HPB xxxx, xxx, xxx
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
5
Table 2 Meta-analysis result of EEN vs. TPN after PD
Outcome
Studies
Participants
Statistical Method
Effect Estimate
P value
I2
Overall postoperative complications
5
669
RR (M-H, Random)
0.95 [0.60, 1.50]
0.84
78%
Delayed gastric emptying
6
721
RR (M-H, Fixed)
0.99 [0.91, 1.08]
0.80
2%
Postoperative hemorrhage
6
726
RR (M-H, Fixed)
0.86 [0.58, 1.27]
0.45
20%
Pancreatic fistula
6
726
RR (M-H, Fixed)
1.33 [0.97, 1.84]
0.08
46%
Mortality
5
689
RR (M-H, Random)
1.88 [0.55, 6.47]
0.32
55%
Length of hospital stay (day)
5
603
MD (IV, Fixed)
−1.46 [−2.04, −0.89]
<0.001
20%
Biliary fistula
4
485
RR (M-H, Fixed)
0.96 [0.40, 2.33]
0.94
0%
Intra-abdominal infection
3
399
RR (M-H, Fixed)
0.62 [0.26, 1.49]
0.28
0%
Wound infection
3
379
RR (M-H, Fixed)
0.53 [0.29, 1.00]
0.05
26%
Lung infection
3
399
RR (M-H, Fixed)
1.03 [0.47, 2.24]
0.94
33%
EEN: Early Enteral Nutrition; TPN, Total Parenteral Nutrition; PD, Pancreaticoduodenectomy; RR, Risk Ratio.
Table 3 Meta-analysis result of EEN + PN vs. EEN/TPN after PD
Outcome
Studies
Participants
Statistical Method
Effect Estimate
P value
I2
Pancreatic fistula
3
531
RR (M-H, Fixed)
1.42 [0.80, 2.50]
0.23
0%
Wound infection
3
531
RR (M-H, Fixed)
0.83 [0.48, 1.44]
0.50
0%
Lung infection
3
531
RR (M-H, Random)
0.95 [0.20, 4.38]
0.95
75%
Mortality
3
531
RR (M-H, Fixed)
0.95 [0.24, 3.83]
0.94
0%
Length of hospital stay (day)
3
531
MD (IV, Random)
5.64 [−1.80, 13.09]
0.14
98%
the article by Gerritsen et al.,19 the standard deviation of continuous variables was not reported and could not contribute to the meta-analysis, which might affect the outcome. In addition, owing to non-availability of the original data on demographic information, disease, and surgical method, subgroup analysis could not be performed, making it unclear whether EEN is potentially more effective or risky in these subgroups. Data of subgroups in the EEN group were also included, such as immune-enhanced EEN and standard EEN or nasojejunal tube and jejunostomy tube. These factors might influence the effect of EEN; however, few studies reported outcomes, making our further research in this regard impossible. In this research, EEN after PD was compared with TPN after PD, and EEN was found to reduce the time and cost of hospitalization. Therefore, EEN might be a more appropriate choice. After combining two nutrition strategies, no obvious advantages or disadvantages were observed, suggesting that such combination was safe and could be used to supplement a single nutritional strategy, especially for EN-intolerant patients.
Authors’ contributions Conceived and designed the experiments: CJ and YG. Literature search and selection: CJ and YG. Analyzed the data: CJ. Contributed reagents/materials/analysis tools: TY, LL and WX. Wrote the paper: CJ and YG. All authors have read and approved the manuscript.
Funding This study was supported by PUMC Youth Fund and the Fundamental Research Funds for the Central Universities [grant number 2017320027] and the Chinese Academy Medical Science Innovation Fund for Medical Students [grant number 20171002-1-16].
Conflicts of interest None. References 1. Schnelldorfer T, Ware AL, Sarr MG, Smyrk TC, Zhang L, Qin R et al. (2008) Long-term survival after pancreatoduodenectomy for pancreatic
Conclusion
adenocarcinoma: is cure possible? Ann Surg 247:456–462. 2. Weimann A, Braga M, Harsanyi L, Laviano A, Ljungqvist O, Soeters P
Compared with TPN, EEN is a safe strategy and can significantly shorten the LOS in patients after PD. For patients with gastrointestinal tolerance, EEN can be considered a priority; otherwise, PN combined with EEN is also a safe strategy.
HPB xxxx, xxx, xxx
et al. (2006) ESPEN guidelines on enteral nutrition: surgery including organ transplantation. Clin Nutr 25:224–244. 3. Weimann A, Braga M, Carli F, Higashiguchi T, Hübner M, Klek S et al. (2017) ESPEN guideline: clinical nutrition in surgery. Clin Nutr 36:623–650.
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002
HPB
6
4. Bozzetti F, Braga M, Gianotti L, Gavazzi C, Mariani L. (2001) Postoperative
17. Stang A. (2010) Critical evaluation of the Newcastle-Ottawa scale for the
enteral versus parenteral nutrition in malnourished patients with gastroin-
assessment of the quality of nonrandomized studies in meta-analyses.
testinal cancer: a randomised multicentre trial. Lancet 358:1487–1492.
Eur J Epidemiol 25:603–605.
5. Mazaki T, Ebisawa K. (2008) Enteral versus parenteral nutrition after
18. Park JS, Chung H, Hwang HK, Kim JK, Yoon DS. (2012) Postoperative
gastrointestinal surgery: a systematic review and meta-analysis of
nutritional effects of early enteral feeding compared with total parental
randomized controlled trials in the English literature. J Gastrointest Surg
nutrition in pancreaticoduodectomy patients: a prosepective, randomized study. J Korean Med Sci 27:261.
12:739–755. 6. Group EC. (2015) The impact of enhanced recovery protocol compli-
19. Gerritsen A, Besselink MG, Cieslak KP, Vriens MR, Steenhagen E, van
ance on elective colorectal cancer resection: results from an Interna-
Hillegersberg R et al. (2012) Efficacy and complications of nasojejunal,
tional registry. Ann Surg 261:1153–1159.
jejunostomy and parenteral feeding after pancreaticoduodenectomy.
7. Visioni A, Shah R, Gabriel E, Attwood K, Kukar M, Nurkin S. (2018)
J Gastrointest Surg 16:1144–1151.
Enhanced recovery after surgery for noncolorectal surgery?: a sys-
20. Guilbaud T, Birnbaum DJ, Loubière S, Bonnet J, Chopinet S, Grégoire E
tematic review and meta-analysis of major abdominal surgery. Ann Surg
et al. (2017) Comparison of different feeding regimes after pancreatoduodenectomy – a retrospective cohort analysis. Nutr J 16.
267:57–65. 8. Gerritsen A, Wennink RA, Besselink MG, van Santvoort HC, Tseng DS,
21. Zhu X. (2013) Effect of early enteral combined with parenteral nutrition in
Steenhagen E et al. (2014) Early oral feeding after pancreatoduodenectomy
patients undergoing pancreaticoduodenectomy. World J Gastroenterol
enhances recovery without increasing morbidity. HPB 16:656–664.
19:5889.
9. Liu C, Du Z, Lou C, Wu C, Yuan Q, Wang J et al. (2011) Enteral nutrition is
22. Lu J, Liu C, Du Z, Liu X, Lv Y, Zhang X. (2016) Early enteral nutrition vs
superior to total parenteral nutrition for pancreatic cancer patients who
parenteral nutrition following pancreaticoduodenectomy: experience
underwent pancreaticoduodenectomy. Asia Pac J Clin Nutr 20:154–160.
from a single center. World J Gastroenterol 22:3821.
10. Gianotti L, Braga M, Gentilini O, Balzano G, Zerbi A, Di Carlo V. (2000)
23. Nagata S, Fukuzawa K, Iwashita Y, Kabashima A, Kinoshita T,
Artificial nutrition after pancreaticoduodenectomy. Pancreas 21:344–351.
Wakasugi K et al. (2009) Comparison of enteral nutrition with combined
11. Seike J, Tangoku A, Yuasa Y, Okitsu H, Kawakami Y, Sumitomo M.
enteral and parenteral nutrition in post-pancreaticoduodenectomy pa-
(2011) The effect of nutritional support on the immune function in the acute postoperative period after esophageal cancer surgery: total parenteral nutrition versus enteral nutrition. J Med Investig 58:75–80. 12. Buscemi S, Damiano G, Palumbo V, Spinelli G, Ficarella S, Monte G et al. (2015) Enteral nutrition in pancreaticoduodenectomy: a literature
tients: a pilot study. Nutr J 8:24. 24. Han H, Pan M, Tao Y, Liu R, Huang Z, Piccolo K et al. (2018) Early enteral nutrition is associated with faster post-esophagectomy recovery in Chinese esophageal cancer patients: a retrospective cohort study. Nutr Cancer 70:221–228. 25. Li B, Liu HY, Guo SH, Sun P, Gong FM, Jia BQ. (2015) Impact of early
review. Nutrients 7:3154–3165. 13. Heyland DK, Montalvo M, MacDonald S, Keefe L, Su XY, Drover JW. (2001) Total parenteral nutrition in the surgical patient: a meta-analysis.
postoperative enteral nutrition on clinical outcomes in patients with gastric cancer. Genet Mol Res 14:7136–7141. 26. Shen Y, Jin W. (2013) Early enteral nutrition after pancreatoduodenec-
Can J Surg 44:102–111. 14. Perinel J, Mariette C, Dousset B, Sielezneff I, Gainant A, Mabrut J et al. (2016) Early enteral versus total parenteral nutrition in patients undergoing pancreaticoduodenectomy. Ann Surg 264:731–737. 15. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP et al. (2009) The PRISMA statement for reporting systematic reviews
tomy: a meta-analysis of randomized controlled trials. Langenbeck’s Arch Surg 398:817–823. 27. Tien Y, Yang C, Wu Y, Hu R, Lee P. (2009) Enteral nutrition and biliopancreatic diversion effectively minimize impacts of gastroparesis after pancreaticoduodenectomy. J Gastrointest Surg 13:929–937.
and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration. BMJ 339:b2700. 16. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ et al. (1996) Assessing the quality of reports of random-
Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10. 1016/j.hpb.2019.06.002.
ized clinical trials: is blinding necessary? Control Clin Trials 17:1–12.
HPB xxxx, xxx, xxx
© 2019 International Hepato-Pancreato-Biliary Association Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article as: Cai J et al., A meta-analysis of the effect of early enteral nutrition versus total parenteral nutrition on patients after pancreaticoduodenectomy, HPB, https://doi.org/10.1016/j.hpb.2019.06.002