Comparison of Biliary Drainage Techniques for MBO: A Meta-Analysis

Abstract

Background:

Malignant biliary obstruction (MBO) impacts patient health and quality of life. Biliary drainage techniques, including percutaneous transhepatic biliary drainage (PTBD) and endoscopic biliary drainage (EBD), are commonly used, yet their relative advantages remain unclear.

Objective:

To compare the effectiveness and complications of EBD and PTBD in treating MBO.

Methods:

A systematic literature search in databases such as PubMed and EMBASE identified case-control trials comparing EBD and PTBD from January 2010 to the present. Five studies comprising 721 participants were analyzed. Outcome measures included technical success rate, incidence of complications, postdrainage pancreatitis, bleeding, major complications, hospital stay duration, and implant metastasis rates. Statistical analysis was conducted using RevMan 5.3 software.

Results:

There was no significant difference between the 2 groups regarding major complications, bleeding incidents, hospital stay, or implantation metastasis rates. However, PTBD showed a significantly lower incidence of complications and postdrainage pancreatitis compared with EBD.

Conclusion:

PTBD may offer advantages over EBD in managing MBO, including fewer complications and reduced incidence of pancreatitis, suggesting PTBD as a potentially safer and more effective treatment option. Nonetheless, further large-scale, high-quality studies are needed to validate these findings.

Malignant biliary obstruction (MBO) is a serious pathological condition caused by malignant tumors such as bile duct cancer, pancreatic cancer, and hilar cholangiocarcinoma, and is characterized by high morbidity and mortality.¹ As these malignant tumors invade or compress the biliary system, bile outflow is blocked, and patients often develop serious complications such as intractable jaundice, recurrent cholangitis, biliary cirrhosis, and even liver failure. Timely and effective relief of biliary obstruction is a key link in improving patient prognosis and quality of life.²

Biliary drainage technology performs a crucial function in the treatment of MBO. It can relieve bile duct hypertension, restore bile drainage, and reduce cholangitis and liver function damage through drainage. At present, the widely used biliary drainage methods in clinical practice mainly include percutaneous transhepatic biliary drainage (PTBD) and endoscopic biliary drainage (EBD). EBD technology mainly achieves biliary drainage through endoscopic retrograde cholangiopancreatography (ERCP). Its advantages are less trauma, better patient tolerance, and other endoscopic treatments can be performed at the same time.³ However, EBD may cause pancreatitis, cholangitis, and other related complications during the operation.⁴ PTBD technology enters the intrahepatic bile duct for drainage through percutaneous puncture. It is suitable for cases where ERCP is unsuccessful or patients whose tumors are difficult to access through endoscopic approaches.⁵ Although PTBD technology is more applicable in some cases, its operation is relatively complicated, and the risk of postoperative bleeding, infection, and biliary implant metastasis is relatively high.

At present, many studies have confirmed that EBD and PTBD are both effective biliary drainage techniques, but in actual clinical applications, the advantages and disadvantages of the 2 are still controversial. The clinical efficacy, complication rate, patient prognosis, and other indicators reported in various studies vary greatly. The existing literature has not yet reached a clear conclusion. The comparison of drainage methods in the treatment of malignant biliary blockage still requires substantiation from rigorous study findings. In this context, further relevant research is very essential. Further reliable scientific research are required to illustrate the clinical efficacy, complication rate, and impact on patient prognosis of the two drainage techniques, so as to provide clinicians with a stronger evidence-based basis when choosing appropriate treatment strategies. Based on this, this study was specially carried out. The present study conducted a systematic, quantitative, and comprehensive analysis of several independent studies of the same nature using meta-analysis. The objective was to evaluate the efficacy and safety of EBD and PTBD drainage techniques in individuals diagnosed with MBO. The findings of this study provide an objective foundation for their clinical implementation and future investigation.

RESEARCH CONTENT AND METHODS

Sources of Literature and Retrieval Methods

Utilize computer search engines such as CNKI, China National Knowledge Infrastructure, Chinese biomedical literature database (CBM), Cochrane Library, PubMed, VIP full-text database, EMBASE, ScienceDirect, and Wanfang Database to conduct searches for pertinent Chinese and foreign journals, conference papers, dissertations, news articles, news articles, and manual retrieval content. Supplement this search with literature tracing techniques. Data were collected from case-control trials in individuals with MBO treated with EBD or PTBD as interventions. The search for literature was done using topic terms and free words. Search terms: clinical efficacy, endoscopic biliary drainage, percutaneous transhepatobiliary drainage, MBO, percutaneous transhepatic biliary drainage, endoscopic biliary drainage, etc. The retrieval time has been since January 2010.

Literature Inclusion and Exclusion Criteria

Literature Inclusion Criteria

(1) Study type: case-control trials of patients with MBO treated with EBD or PTBD as intervention measures at home and abroad. (2) Study subjects: Individuals with a clear diagnosis of MBO, with the diagnostic criteria referring to relevant literature,² with no restrictions on age, sex, or nationality. (3) Intervention measures: EBD and PTBD were used as biliary drainage methods. (4) Reported one or more of the following outcome indicators: (1) technical success rate; (2) incidence of postdrainage complications; (3) incidence of postdrainage pancreatitis; (4) incidence of postdrainage bleeding; (5) incidence of major postdrainage complications; (6) length of hospital stay; (7) implant metastasis rate, etc.

Literature Exclusion Criteria

(1) It is not a case-control study; (2) incomplete data report renders the data unusable; (3) the scientific information is repeated, and the most recent study is cited; (4) the assessment of the research effectiveness lies insignificant; (5) clinical cases; (6) review-related literature.

Quality Assessment and Data Extraction

(1) The Newcastle-Ottawa Scale (NOS) tool was utilized to assess the quality of observational studies. Studies with a NOS score of ≥6 were categorized as medium to high quality, whereas those with an NOS score <6 were classified as low quality. (2) Literature screening and data extraction: 2 researchers conducted separate screenings of the literature, extracted data, assessed the quality of the material, and verified the information obtained. Should any dispute arise, it was either settled via conversation or a third researcher was enlisted to aid in the decision-making process. Excel office software and NoteExpress literature management software were accustomed to organize and extract the study data. In cases where the data from the literature presented was insufficient, the author of the manuscript was approached to request further information. The data extraction content included (1) basic information: author, publication time, number of cases; (2) intervention measures: regimen, course of treatment; (3) outcome indicators: technical success rate, clinical success rate, treatment crossover rate, incidence of postdrainage complications, incidence of postdrainage pancreatitis, incidence of postdrainage bleeding, incidence of major postdrainage complications, length of hospital stay, and transplant transfer rate.

Statistical Processing

This Meta-analysis was carried out utilizing RevMan5.3 software, which was developed by the Cochrane Collaboration. Inputted into RevMan5.4 for analysis were the counting data of the PTBD group and the EBD group, which included the technical success rate, clinical success rate, treatment crossover rate, incidence of complications after drainage, incidence of pancreatitis after drainage, incidence of bleeding after drainage, incidence of major complications after drainage, and implantation metastasis. The effect measure used was the relative risk ratio (OR). Data on the average and variability of the hospitalization duration for both the EBD group and the PTBD group were entered into RevMan5.4 for statistical analysis. As the effect indicator, the weighted mean difference (WMD) was used to calculate a 95% confidence interval (95% CI). First, the existence of heterogeneity among the studies was determined using the χ² test. Studies included in the meta-analysis are considered homogenous if the interquartile range (I ²) is <50% and the significance level (P) is more than 0.1. The modified effect model may be used in the meta-analysis in these circumstances. The random effects model is used if the combined effect is necessary to evaluate the homogeneity of the included studies, the significance level (P) is <0.1, and the I ² is <50%. If the heterogeneity’s cause cannot be found and the P-value is <0.1, descriptive analysis is used in its place rather than a meta-analysis. The included literature’s publication bias is examined using an inverted funnel plot. To evaluate the funnel plot’s asymmetry, use the Egger test. The Trim and Fill method may be used to adjust the effect size of publication bias and correct the funnel plot if the test’s P-value is <0.1.

RESULTS AND ANALYSIS

Literature Search Results and Basic Information of Included Literature

The literature was examined and analyzed after the standards of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA). Of the 1631 papers acquired by computer database retrieval, 1205 articles were retained after eliminating duplicate research. Following a preliminary review of the publication titles and abstracts, 853 items were found. Subsequently, the whole texts were thoroughly reviewed. A total of 848 papers without comprehensive data and key outcome indicators were excluded. The final analysis included 5 clinical control trials, totaling 721 samples. The flow chart for literature screening is seen in Figure 1. Table 1 displays the essential qualities of the literature that was gathered.

FIGURE 1.

The PRISMA flow diagram for literature screening.

TABLE 1.

Basic Characteristics of the Literature

		Sample size			Interventions
Included literature	Year of publication	C	T	Study design	C	T	Outcome measures
Ba et al6	2020	99	81	Cohort study	EBD	PTBD	②③④⑤⑥
Jo et al7	2016	61	37	Cohort study	EBD	PTBD	①②③④⑤⑦
Kim et al8	2015	44	62	Cohort study	EBD	PTBD	①②③④⑤⑥⑦
Zhang et al9	2017	92	104	Cohort study	EBD	PTBD	①⑥
Hirano et al10	2014	67	74	Cohort study	EBD	PTBD	②③⑤⑥

Note: ① Technical success rate; ② the incidence of complications after drainage; ③ the incidence of postdrainage pancreatitis; ④ the incidence of bleeding after drainage; ⑤ the incidence of major complications after drainage; ⑥ length of hospital stay; ⑦ implant transfer rate.

Assessment of Methodological Quality of Included Articles

All reviewed studies provided detailed descriptions of intervention methods and observation indicators. However, none of the studies thoroughly reported the number and reasons for blinding procedures, missed follow-ups, or withdrawals. According to the NOS scale assessment, studies with a score of <6 were classified as low quality, while those scoring ≥6 were considered high quality (Table 2).

TABLE 2.

Literature Quality

Study	Representativeness of the exposed cohort	Selection of the nonexposed cohort	Ascertainment of exposure	Demonstration that the outcome of interest was not present at the start of the study	Comparability of cohorts on the basis of the design or analysis	Assessment of outcome	Was the follow-up long enough for outcomes to occur	Adequacy of follow-up of cohorts	Quality score
Ba Y6	★	★	★	★	★★	★	★	★	9
Jo JH7	★	★	★	★	★	★	★	★	8
Kim KM8	★	★	★	★	★	★	★	★	8
Zhang XF9	★	★	★	★	★★	★	★	★	9
Hirano S10	★	★	★	★	★★	★	★	★	9

“★” = 1 point of the Newcastle-Ottawa scale; “★★”=2 point of the Newcastle-Ottawa scale; studies get 1 point at each category once they meet the criteria.

Meta-Analysis Results

Technical Success Rate

Comprising a total of 721 samples, this work included 5 case-control trials.^6–10 A rigorous meta-analysis was completed to assess the technical success rates of the 2 groups. The heterogeneity test findings indicated that χ²=1.38, df=2, P=0.50, I ²=0%, suggesting the absence of significant heterogeneity among the research data analyzed. According to Figure 2 ’s fixed effect model analysis, there was no considerable change in the 2 groups’ technical success rates (P>0.05).

FIGURE 2.

Forest analysis diagram comparing the technical success rates of the 2 groups. df indicates degrees of freedom; I ², Higgins’ I ² statistic; Z, Z-score; χ², χ² test.

The Incidence of Complications After Drainage

A systematic review was conducted to analyze the incidence of complications after drainage in both groups. The heterogeneity test revealed that the variables χ²=6.50, df=3, P=0.09, and I ²=54% indicated significant heterogeneity among the study data considered. Analyzed using the random effects model, Figure 3 revealed a considerable change (P<0.05) in the incidence of complications after drainage between the PTBD group and the EBD group.

FIGURE 3.

Forest analysis chart comparing the incidence of complications after drainage between the 2 groups. df indicates degrees of freedom; I ² Higgins’ I ² statistic; Z, Z-score; χ², χ² test.

The Incidence of Postdrainage Pancreatitis

A systematic review and statistical analysis were carried out to ascertain the occurrence of postdrainage pancreatitis in the 2 specified groups. The sample heterogeneity test yielded χ²=0.76, df=3, P=0.86, I ²=0%, suggesting the absence of significant heterogeneity among the study data collected. As seen in Figure 4, the fixed effect model analysis revealed a considerable change (P<0.05) in the incidence of postdrainage pancreatitis between the PTBD group and the EBD group.

FIGURE 4.

Forest analysis chart comparing the incidence of pancreatitis after drainage between the 2 groups. df indicates degrees of freedom; I ², Higgins’ I ² statistic; M-H, Mantel-Haenszel; Z, Z-score; χ², χ² test.

The Incidence of Bleeding After Drainage

A comparative meta-analysis was carried out to examine the occurrence of bleeding after drainage in the 2 groups. The sample heterogeneity test yielded χ²=1.21, df=1, P=027, I ²=18%, suggesting the absence of considerable heterogeneity among the study data collected. Statistical analysis using the fixed effect model revealed no considerable change in the occurrence of bleeding after drainage involving the 2 groupings (P>0.05), as seen in Figure 5.

FIGURE 5.

Forest analysis chart comparing the incidence of bleeding after drainage between the 2 groups. df indicates degrees of freedom; I ², Higgins’ I ² statistic; M-H: Mantel-Haenszel; Z, Z-score; χ², χ² test.

The Incidence of Major Complications After Drainage

A meta-analysis took place to examine the occurrence of significant problems after drainage in the 2 groups. Based on the findings of the heterogeneity test, χ²=2.29, df=2, P=0.32, I ²=13%, it is evident that there was no significant heterogeneity among the study data included. The fixed effect model revealed that the occurrence of significant problems after drainage was greater in the EBD group, but there was no considerable change (P>0.05), as seen in Figure 6.

FIGURE 6.

Forest analysis chart comparing the incidence of major complications after drainage involving the 2 groupings. df indicates degrees of freedom; I ², Higgins’ I ² statistic; M-H, Mantel-Haenszel; Z, Z-score; χ², χ² test.

Length of Hospital Stay

A meta-analysis was completed to examine the duration of hospitalisation in the 2 groups. The heterogeneity test revealed that the parameters χ²=5.15, df=2, P<0.08, and I ²=61% indicated significant heterogeneity among the study data contained. The random response model analysis showed no statistically significant reduction in hospitalization time for the PTBD group in contrast to the TBD group (P>0.05, Fig. 7).

FIGURE 7.

Forest analysis chart comparing the length of hospital stay between the 2 groups. df indicates degrees of freedom; I ², Higgins’ I ² statistic; Z, Z-score; χ², χ² test.

Implant Transfer Rate

A meta-analysis was completed to compare the rates of implant metastases in the 2 groups (Fig. 8). No considerable change was seen in the frequencies of implant metastases involving the 2 groupings (P>0.05).

FIGURE 8.

Forest analysis chart comparing the planting transfer rates of the 2 groups. df indicates degrees of freedom; I ², Higgins’ I ² statistic; M-H, Mantel-Haenszel; Z, Z-score; χ², χ² test.

Publication Bias Analysis

Funnel plots were drawn based on the technical success rate, incidence of postdrainage complications, incidence of postdrainage pancreatitis, incidence of postdrainage bleeding, incidence of major postdrainage complications, hospitalization time, and metastasis implantation rate of the 2 groups of patients, and publication bias analysis was performed (Figs. 9–15). The findings indicated that the majority of the funnel plots exhibited symmetry, while a minority displayed asymmetry. This observation implies the presence of a publication bias within the literature included, maybe attributable to the heterogeneity of the study, as well as the small number of publications that are featured.

FIGURE 9.

Funnel chart based on technical success rate.

FIGURE 15.

Funnel diagram based on the planting transfer rate.

FIGURE 10.

Funnel plot based on the incidence of complications after drainage.

FIGURE 11.

Funnel plot based on the incidence of pancreatitis after drainage.

FIGURE 12.

Funnel plot based on the incidence of bleeding after drainage.

FIGURE 13.

Funnel plot based on the incidence of major complications after drainage.

FIGURE 14.

Funnel plot based on hospital stay.

DISCUSSION

MBO is a serious pathological condition in the biliary system, usually caused by cholangiocarcinoma, pancreatic cancer, or other malignant tumors, which leads to obstruction of bile excretion, leading to symptoms such as jaundice and liver function damage. According to statistics,¹¹ as a result of population aging and rising cancer incidence, the number of cancer cases is increasing year by year. MBO not only seriously affects the individual’s quality of life, but may also significantly shorten the individual’s survival due to delayed treatment. Therefore, finding effective treatment options is of great clinical significance to improve patient prognosis.

Surgical intervention continues to be the most efficient therapy for respectable malignant tumors affecting the liver and pancreas.^12–14 Ensuring the successful outcome of surgery for malignant hepatobiliary and pancreatic tumors with obstructive jaundice requires the effective relief of biliary blockage and reduction of hyperbilirubinemia. Nevertheless, a growing body of research has shown that PBD not only does not provide any advantages in terms of survival for individuals with malignant biliary problem, but also extends the duration of hospital stay and raises the occurrence of postoperative problems.¹⁵ Even so, guidelines from the United States,¹² Europe,¹³ Japan,^16–18 and other countries highly suggest suitable PBD for people with MOJ. Indeed, there is much debate among the global hepatobiliary and pancreatic community over the selection of the most suitable draining technique: Europe and America¹² standards advocate for PTBD; however, Japanese guidelines firmly endorse EBD.¹⁸ In contrast to EBD, PTBD is characterized by its simplicity of operation and simpler promotion. A total of 721 samples from 5 clinical control studies were used in this research. Meta-analysis of the technical success rate of the 2 drainage techniques suggested that no considerable change was seen in the technical success rate involving the 2 groupings, suggesting that these methods are similar in technical success rate. Nevertheless, PTBD may has certain advantages in technical operation, as Jo et al⁷ reported the crossover treatment rate of 2 patients (4.7%) in the PTBD group was lower than that of 25 patients (15.4%) in the EBD group. Besides, in a randomised controlled trial,¹⁹ 15 patients (56%) in the EBD group required additional PTBD due to technical failure, treatment failure, or complications, while in the PTBD group, only 1 patient (4%) received additional EBD.

Our analysis compared the incidence of postdrainage complications, including pancreatitis, bleeding, and major complications, between PTBD and EBD. The results indicated that while EBD had a higher incidence of postdrainage complications and pancreatitis, no significant differences were observed in major complications or postdrainage bleeding between the 2 methods. This aligns with the findings of Al Mabjoub et al,²⁰ suggesting that EBD, despite being less invasive, carries a higher risk due to its reliance on endoscopic techniques, which may increase procedural complexity and the likelihood of complications such as pancreatitis and bleeding.²¹ Operator experience is also a critical factor, as EBD requires skilled endoscopists to minimize risks.²² Beyond complication rates, clinical decision-making should also consider the patient's economic burden. PTBD, while associated with lower crossover treatment rates, may cause inconvenience and additional costs due to the long-term indwelling catheter.^23,24 Conversely, EBD, despite its higher risks, is easier to perform and may facilitate faster postoperative recovery. Therefore, selecting the appropriate drainage method should involve a comprehensive evaluation of patient-specific factors, including tumor location, biliary anatomy, physical condition, anticipated surgical risks, and economic constraints. Preoperative imaging and laboratory tests should be utilized to assess biliary obstruction severity and liver function to tailor individualized treatment plans. In addition, our meta-analysis did not quantitatively evaluate long-term prognosis (eg, overall survival, recurrence-free survival) due to heterogeneity in data presentation among included studies. Existing literature presents conflicting results: Hirano et al¹⁰ suggested that EBD might provide a better prognosis by reducing peritoneal seeding, whereas Zhang et al⁹ found no significant differences in survival outcomes between EBD and PTBD. Standardizing long-term outcome reporting in future studies would facilitate more robust comparisons and evidence-based decision-making.

Implantation metastasis is a known dissemination mechanism in abdominal malignancies, occurring via direct or percutaneous routes through blood or lymphatic spread. Hepatobiliary and pancreatic tumors, especially hilar cholangiocarcinoma, are highly susceptible to this process. Studies indicate that implantation metastases can develop even in the absence of preoperative biliary drainage, with a prevalence of up to 15.9%.²⁵ Our analysis found no significant difference in hospitalization duration or implantation metastasis rates between PTBD and EBD. However, implantation risk depends on multiple factors, including tumor biology, immune status, and surgical precision. Therefore, in addition to selecting an optimal drainage method, meticulous intraoperative strategies are essential to minimize metastasis risk. Economic burden is another crucial factor in clinical decision-making. While our meta-analysis primarily assessed clinical outcomes, some studies have addressed cost-related aspects such as hospital stay duration and reintervention rates. PTBD was linked to a shorter hospital stay, potentially reducing inpatient costs.⁸ However, EBD, despite sometimes requiring longer hospitalization, offers advantages such as lower invasiveness and fewer external catheter-related complications.⁷ Conversely, EBD has been associated with higher rates of postdrainage pancreatitis,⁶ increasing health care costs due to additional treatments. Reintervention rates also vary: PTBD generally requires fewer secondary procedures, whereas EBD has a higher incidence of stent occlusion, necessitating repeat interventions.¹⁰ In addition, PTBD carries a risk of catheter tract seeding, which could further increase the economic burden.⁹ Considering these factors, the choice of biliary drainage should be individualized, weighing both clinical and economic implications to optimize patient outcomes.

Implantation metastasis is a known mechanism of distant spread in abdominal malignancies, occurring through direct or percutaneous dissemination via blood or lymphatic routes. It can manifest as thoracic, peritoneal, or body wall implantation. Due to their anatomical location, hepatobiliary and pancreatic tumors, particularly hilar cholangiocarcinoma, are highly susceptible to this process. Studies indicate that implantation metastases can occur in malignant obstructive jaundice patients even without preoperative biliary drainage, with a reported prevalence of up to 15.9%.²⁵ Our analysis found no significant difference in hospitalization duration or implantation metastasis rates between PTBD and EBD. However, metastasis risk depends on tumor biology, immune status, and surgical precision. Therefore, beyond selecting an optimal drainage method, meticulous intraoperative techniques are essential to minimize implantation risk.

All in all, PTBD has demonstrated good safety and efficacy in managing MBO, particularly in reducing complications and pancreatitis incidence. However, both PTBD and EBD have distinct indications and limitations, necessitating individualized clinical decision-making. This study has certain limitations. The limited number of clinical trials and modest sample size may affect the generalizability of the findings. In addition, variations in tumor types, patient conditions, and surgical approaches across studies introduce potential bias. Moreover, the lack of long-term follow-up restricts a comprehensive evaluation of long-term prognosis. Future research should focus on expanding sample sizes, conducting multicenter randomized controlled trials, and ensuring extended follow-up to provide more definitive conclusions.