Effect of preoperative biliary drainage in patients with moderately severe obstructive jaundice on the outcomes of pancreatoduodenectomy: A meta-analysis study


 Background The purpose of this meta-analysis was to examine the impact of preoperative biliary drainage (PBD) on the perioperative outcomes of pancreatoduodenectomy (PD) in patients with total bilirubin >100 umol/L.Methods In this meta-analysis, studies that compared the perioperative outcomes of PBD and non-PBD patients with total bilirubin >100 umol/L, and were published in EMBASE, PubMed, the Cochrane library, Web of Science, VIP database, Wanfang data, Chinese biomedical literature and CNKI database from inception up to October 2019 were included. The odds ratios (OR) or mean differences were calculated with 95% confidence intervals (CI).Results Nine trials with 744 patients, which compared PBD (267 patients) with non-PBD (477 patients), were included. There was no significant difference in perioperative mortality between these two groups (OR: 0.51, 95% CI: 0.19 to 1.39; P =0.19). Postoperative hospital stay (mean difference: -2.35, 95% CI: -3.70 to -1.00; P =0.0007), operating time (mean difference: -33.03, 95% CI: -44.14 to 21.93; P <0.00001), estimated blood loss (mean difference: -141.18, 95% CI: -213.25 to -69.11; P =0.0001) and overall morbidity (OR: 0.68, CI: 0.48 to 0.95; P =0.02) were significantly lower in the PBD group than in the non-PBD group.Conclusion Patients who received PBD had similar perioperative mortality, but had decreased postoperative hospital stay, operating time, estimated blood loss and overall morbidity, when compared to patients without PBD. Therefore, PBD should be routinely performed for patients planned for PD with a total bilirubin of >100 umol/L.


Background
Malignant obstructive jaundice occurs due to luminal obstruction of the extrahepatic bile duct by tumors, such as cholangiocarcinoma, pancreatic cancer and periampullary carcinoma. Surgical excision remains as the only curative treatment for such cases. Studies have shown that obstructive jaundice can lead to liver metabolic dysfunction 1 , renal failure 2 and decreased immune functions 3 , thereby resulting in higher postoperative complications. Shaib 4 et al. studied 2,306 patients with obstructive jaundice, and found that preoperative biliary drainage (PBD) was helpful in reducing postoperative morbidities. However, McPhern 5 et al. reported that the mortality of obstructive jaundice with PBD was distinctly higher than that of non-PBD patients. To date, the indications and criteria for PBD in obstructive jaundice patients who require surgery remains unclear.
The present study aims to explore the impact of PBD in patients with moderately severe obstructive jaundice (total bilirubin >100 umol/L) on perioperative outcomes of pancreatoduodenectomy (PD).

Search strategy
Relevant studies were retrieved from the databases, including EMBASE, PubMed, the Cochrane library, Web of Science, VIP database, Wanfang Data, Chinese biomedical literature and CNKI database, from inception up to October 2019, using the following search terms: "preoperative biliary drainage", "obstructive jaundice" and "pancreaticoduodenectomy". Then, a systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA) 6 and Cochrane Collaboration 7 guidelines.

Selection criteria
Studies that satis ed the following criteria were included in the present meta-analysis: original articles reported in the English or Chinese language; studies on patients with obstructive jaundice (total bilirubin >100 umol/L), who underwent pancreaticoduodenectomy; studies with PBD and non-PBD groups; studies that reported the perioperative mortality, hospital stay, operating time, estimated blood loss and overall morbidity. The exclusion criteria were as follows: studies with irretrievable data; non-comparative studies; letter to editors, comments, or meta-analyses; studies with duplication of data; studies with <10 patients in either group.

Data extraction
Two investigators independently extracted the data. All discrepancies were solved by consensus, or with the help of a third investigator. The following information were extracted from each study: study groups, author, year of publication, total bilirubin, number of patients, perioperative mortality, hospital stay, operating time, estimated blood loss and overall morbidity.

Statistical analysis
The present meta-analysis was conducted according to the guidelines of the Cochrane Handbook for Systematic Reviews of Interventions 5.1.0. The observed indexes were expressed as an odds ratio (OR) with 95% con dence interval (CI) for dichotomous variables, and the mean difference was expressed with 95% CI for continuous variables. The Mantel-Haenszel method was applied to combine the OR, and the inverse variance was used to merge the mean difference for the observed indexes. The heterogeneity among studies was evaluated using a standard X 2 -test with the standard of signi cance set at a Pvalue of <0.05, and assessed via the I 2 statistic with the standard of signi cance set at I 2 >30%. If heterogeneity was present, a random effects model was applied for the meta-analysis. If heterogeneity was absent, a xed effects model was used. The OR or mean difference with 95% CI, and the heterogeneity of either of the subgroups were respectively calculated. The subgroup differences were assessed, and P<0.05 was considered representative of statistical signi cance. All statistical analyses were carried out using Review Manager Version 5.3. For assessing the quality of these studies, bias exploration using a funnel plot and Egger's regression method was planned, but this could not be performed due to the small number of trials included in the present review.

Results
After the initial search in all these databases, a total of 881 studies were retrieved. After the preliminary screening and elimination of articles with duplication of data, 89 articles were selected. After reading the full text of these articles and excluding articles from which data could not be extracted, nine articles (Table   1, Fig. 1) were selected [8][9][10][11][12][13][14][15][16] .

Sensitivity analysis
Since these prospective and retrospective studies were analyzed for mortality, postoperative hospital stay and overall morbidity, a sensitivity analysis to subtotal the plots by prospective studies vs. retrospective studies was performed. There were no signi cant differences in OR or mean difference between these prospective and retrospective studies (all, P>0.05; Figs. 3, 5, 7, 9 and 11).

Discussion
Studies conducted in the end of the 20 th century revealed that upfront surgery in patients with obstructive jaundice is associated with increased postoperative complications and mortality. 18,19 Hence, PBD became a routine procedure for such patients. However, subsequent studies in the 21 st century revealed that PBD itself leads to various complications, especially infectious complications 20 . The positive rate of bile culture after an operation was also found to increase 21 . In order to resolve this controversy, this meta-analysis was conducted to determine the impact of PBD in patients with total bilirubin >100 umol/L, in terms of postoperative mortality, hospital stay, operating time, estimated blood loss and overall morbidity.
The main causes of postoperative death in patients with obstructive jaundice are infection and hemorrhage. Studies have shown that high bilirubin levels decrease the body'' immunity and increase the risk of infection 22,23 . PBD can effectively reduce the serum bilirubin levels, and help restore appetite, reduce symptoms and improve the immune status. However, in the present study, no statistical difference in perioperative mortality was found between the PBD and non-PBD groups (P=0. 19). This nding may have been due to the small sample size of the included studies. However, a previous meta-analysis conducted by Moole et al. on 3,532 patients also revealed no reduction in mortality by PBD. 19 In the present meta-analysis, it was found PBD is associated with signi cantly lower operative time and estimated blood loss. This nding was probably because the high bilirubin levels in the body can damage the capillary cells and affect the coagulation pro le. 24 This causes di culty in achieving hemostasis during the operation. Although operation time and intraoperative blood loss are more closely correlated to the surgical expertise and extent of disease, taking measures to reduce the level of serum bilirubin before surgery can reduce operative di culty to a certain extent.
In the present study, lower incidence of postoperative complications was found in the PBD group. Although the occurrence of complications was correlated to the surgeon's experience, stage of the disease and other patient factors, such as co-morbidities 25 , the reduction in serum bilirubin levels may be bene cial for reducing the incidence of anastomotic leakage 26 . The decrease in bilirubin levels also helps the body overcome the surgical stress. 27 Although the total length of hospital stay of patients with PBD was slightly higher, when compared to non-PBD patients, the present study revealed that the postoperative hospital stay in the PBD group was signi cantly shorter than that in the non-PBD group. This observation was probably correlated to the decrease in postoperative complications and faster recovery. Studies have shown that in order to reduce the preoperative hospital stay in patients who received PBD, after a successful PBD, these patients can be encouraged to go home, improve their oral intake, and undergo regular follow-ups of liver functions. Once the serum bilirubin level normalizes, these patients can be readmitted for surgery. 28,29 There were some limitations in the present study. First, due to the strict selection criteria, many studies with a larger sample size could not be included.
Second, the infectious and non-infectious complications could not be separately compared between the PBD and non-PBD groups.

Conclusions
In conclusion, moderately severe obstructive jaundice patients should receive preoperative biliary drainage. Preoperative biliary drainage can effectively shorten the postoperative hospital stay and operating time, and reduce intraoperative blood loss and complications. Abbreviations preoperative biliary drainage (PBD); pancreatoduodenectomy (PD); odds ratios (OR); con dence intervals (CI) Declarations Ethics approval and consent to participate: Not applicable. Consent for publication: Not applicable.
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Competing interests: The authors declare that they have no competing interests.
Funding: Not applicable.
The owchart of the selection process for identifying relevant articles for the meta-analysis  A forest plot of the individual data and pooled estimate of postoperative hospital stay in the included studies. The inverse variance xed effect model was applied. The mean difference was presented with 95% CI.

Figure 6
A forest plot of the individual data and pooled estimate of operative time in the included studies. The inverse variance xed-effect model was applied. The mean difference was presented with 95% CI.

Figure 8
A forest plot of the individual data and pooled values of estimated blood loss in studies with PBD and non-PBD groups. The inverse variance xed effects model was applied. The mean difference was presented with 95% CI.

Figure 10
A forest plot of the individual study proportions and pooled estimate of overall morbidities in the PBD and non-PBD groups. A Mantel-Haenszel xed effects model was applied. The odds ratios were presented with 95% CI.