Safety and efficacy of three-dimensional reconstruction technology-assisted percutaneous transhepatic biliary drainage: A meta-analysis

Abstract

BACKGROUND

Percutaneous transhepatic biliary drainage (PTBD) is one of the primary clinical treatment options for patients with obstructive jaundice. In recent years, PTBD assisted by three-dimensional (3D) reconstruction technology has been widely implemented, but its advantages over traditional methods remains inconclusive. Thus, a discussion is warranted.

AIM

To explore the safety and efficacy of 3D reconstruction technology-assisted PTBD.

METHODS

We systematically searched the databases including the Cochrane Library, PubMed, EMBASE, Web of Science and China National Knowledge Infrastructure. The search period extended from the establishment of each database to November, 2024. We screened the literature according to predefined inclusion and exclusion criteria, assessed the quality of the studies, and extracted data. Meta-analysis was performed using Revman 5.4.1 software.

RESULTS

A total of 15 studies were included, involving 1434 patients. The results of the meta-analysis showed that compared with the traditional group, the overall post-operative complications rate in the 3D reconstruction technology group was significantly lower [odds ratio = 0.25; 95% confidence interval (CI): 0.17-0.36, P < 0.00001]. The overall puncture success rate in the 3D reconstruction group was better than those in the traditional group (odds ratio = 3.61; 95%CI: 1.98-6.55, P < 0.0001). However, there was no significant difference between the two groups in the reduction levels of postoperative total bilirubin (mean difference = -1.38; 95%CI: -3.29 to 0.53, P = 0.16). Subgroup analysis were conducted on the surgery time according to guidance stages of the 3D reconstruction, 3D reconstruction imaging modalities, and types of studies. The results were stable, with no significant changes observed.

CONCLUSION

3D reconstruction technology significantly improves the puncture success rate and safety of PTBD. However, it has no significant advantage in bile drainage effectiveness. Continued research is warranted to further explore its clinical value and optimize its application.

Key Words: Three-dimensional reconstruction technology; Percutaneous transhepatic biliary drainage; Obstructive jaundice; Complications; Liver function

Core Tip: This meta-analysis of 15 studies demonstrates that three-dimensional reconstruction technology significantly improves the safety and puncture success rate of percutaneous transhepatic biliary drainage compared to traditional methods. Although no significant advantage in bile drainage efficacy was observed, the technology exhibits critical clinical value by reducing complication risks through precise preoperative planning and real-time intraoperative navigation. Future large-scale studies, particularly stratified studies based on biliary anatomy and disease subtypes, are required to clarify its applicability across diverse clinical scenarios, thereby advancing standardized protocols and widespread clinical adoption of this innovative approach.

INTRODUCTION

The accumulation of bile not only increases the pressure of intrahepatic and extrahepatic biliary tract, causing liver function damage, but also excessive bilirubin entering the bloodstream can affect the normal function of other organs, and even cause organ failure[1]. Research has shown that actively reducing jaundice and drainage is of great significance in improving the prognosis of patients with obstructive jaundice[2].

Since Glenn et al[3] first reported percutaneous transhepatic biliary drainage (PTBD) in 1962, this procedure has continued to mature with the advancement of modern technology. Due to its low anesthesia requirements, convenient operation, and flat learning curve, it has become the preferred clinical treatment, especially for end-stage patients in poor general condition who cannot tolerate general anesthesia[4]. Accurate positioning, precise guidance, and effective drainage are crucial for PTBD. However, traditional methods are limited in these aspects[5]. Studies have indicated that the risk of postoperative complications is higher with traditional methods, particularly in terms of postoperative bleeding and biliary fistula, etc[6].

Three-dimensional (3D) reconstruction technology involves reprocessing traditional imaging data to create individualized models. Based on these models, it enables preoperative localization and simulation, as well as real-time intraoperative monitoring and navigation, etc[7,8]. In recent years, 3D reconstruction technology has been widely applied in PTBD due to its advantages of multi-angle, intuitive, scientific, and precise imaging. However, whether 3D reconstruction technology is significantly superior to traditional methods in clinical practice remains unclear. Current controversies primarily focus on three key aspects: (1) Whether it can significantly improve puncture success rate; (2) Whether it can effectively reduce post-operative complication rates; and (3) Whether it can enhance biliary drainage efficacy. Notably, existing studies demonstrate marked heterogeneity in aspects such as inclusion criteria, statistical methods, and sample sizes. Additionally, there is a lack of high-quality evidence derived from meta-analysis. Therefore, this study aims to synthesize the existing evidence through a meta-analysis, provide a reference framework for clinical decision-making, and offer guidance for future research directions.

MATERIALS AND METHODS

Search strategy

Two independent researchers conducted a systematic literature search in PubMed, EMBASE, Cochrane Library and China National Knowledge Infrastructure. The search period ranged from the inception of each database to November 7, 2024. A hybrid strategy combining subject headings and free-text terms was employed. Search terms mainly included “three-dimensional reconstruction”, “three-dimensional visualization”, and “percutaneous transhepatic biliary drainage (PTBD)”. The search strategy was dynamically adjusted according to the indexing systems of each database, using synonym expansions. The search was limited to human studies, with no language restrictions, and reference lists of literature were also manually examined for eligible studies to ensure the recall.

Selection criteria

The study inclusion criteria are as follows: (1) Study participants were patients diagnosed with obstructive jaundice; (2) The study compared the use of 3D reconstruction technology-assisted PTBD with traditional methods; and (3) The literature provided extractable or calculable data on at least one of the following: Surgery time, number of punctures, incidence of specific postoperative complications, overall post-operative complications rate, or reduction levels of postoperative total bilirubin, aspartate transaminase, and alanine transaminase. The exclusion criteria are as follows: (1) Articles published as abstracts, reviews, case reports, meta-analysis, letters, or conference proceedings; (2) Studies with unavailable or incomplete data, or lacking surgical parameters, postoperative complications, and follow-up results; (3) Non-human studies; (4) Uncontrolled single-group analyses; and (5) Duplicate publications.

Data extraction

Two researchers independently screened the literature according to the inclusion and exclusion criteria, extracted data, and conducted data verification and quality assessment. if necessary, adjudicated by a third researcher. Extracted data includes: First author, year of publication, patient demographics (e.g., race, age), sample size, group allocation ratio (3D reconstruction vs traditional), surgery time, number of punctures, incidences of postoperative bleeding, biliary fistula, infection, overall post-operative complication rate, and reduction levels of postoperative total bilirubin, aspartate transaminase and alanine transaminase, along with other relevant outcomes.

Quality assessment

The quality of included randomized controlled trials was conducted using the Cochrane Risk of Bias Assessment Tool. For cohort studies, the Newcastle-Ottawa Scale was applied[9], with a total score of 9 points. A score > 6 was considered indicative of high-quality research. Two independent reviewers assessed study quality, resolving any disagreements through discussion or third-party adjudication.

Statistical analysis

Meta-analysis was performed using Review Manager 5.4.1. For continuous variables, such as surgery time, number of punctures, reduction levels of postoperative total bilirubin, aspartate transaminase and alanine transaminase, the mean difference (MD) and 95% confidence interval (CI) were calculated. For dichotomous variables such as incidence rate of postoperative complications and overall complications rate, odds ratios (ORs) with 95%CIs were used. Heterogeneity among studies was measured using the χ² test and quantified by the I² statistic. If I² ≥ 50% or P < 0.10, significant heterogeneity was assumed, and a random-effects model was applied. Otherwise, a fixed-effect model was used. Funnel plots were employed to assess publication bias. A P < 0.05 was considered statistically significant.

RESULTS

Literature search results

As shown in Figure 1, a total of 4003 studies were identified through the initial search strategy, and an additional 4 studies were determined by screening the references of relevant studies. After removing duplicates, 1520 studies remained. Of these, 1352 were excluded based on title and abstract screening. Following full-text review of 168 articles, 153 were excluded, a total of 15 studies were finally included in the meta-analysis.

Figure 1  Flow chart of article screening.

Study characteristics and quality assessment of included studies

As shown in Table 1, the final analysis included 15 studies involving 1434 patients. These comprised 8 randomized controlled trials and 7 cohort studies, all published between 2016 and 2024. Three studies were obtained from English-language databases, while twelve were from Chinese databases. Among the total patients, 742 patients received PTBD assisted by 3D reconstruction technology, and 692 received traditional methods. The quality of randomized controlled trials was assessed using the Cochrane Collaboration's risk of bias assessment tool (Figure 2), and cohort studies were evaluated with the Newcastle-Ottawa Scale (Table 2).

Figure 2  Article quality evaluation chart: Randomized controlled trials.

Table 1 Characteristics of the 15 included studies.

Ref.	Sample size 3D/tradition	Age 3D/tradition, mean ± SD	Gender (male/female) 3D tradition	Timing of 3D	Imaging modality	Design
Liu et al[22], 2023	81/73	66.5 ± 10.6/67.9 ± 11.4	46/35, 38/35	Pre-operation and intraoperation	3D of CT and MR	Cohort study
Li et al[10], 2020	42/45	NR	NR	Intraoperation	3D of CT	RCT
Zhou et al[11], 2021	30/30	61.1 ± 5.3/60.3 ± 5.5	20/10, 18/12	Pre-operation	3D of MR	Cohort study
Li et al[23], 2017	63/64	38.73 ± 11.34/35.16 ± 10.65	34/29, 32/32	Intraoperation	3D of ultrasound	Cohort study
Duan et al[12], 2017	50/50	47.6 ± 6.7/46.5 ± 5.4	28/22, 27/23	Intraoperation	3D of ultrasound	RCT
Zhuo et al[13], 2024	32/32	NR	22/10, 21/11	Intraoperation	3D of ultrasound	RCT
Zhou et al[24], 2022	120/60	NR	NR	Pre-operation	3D of CT and MR	Cohort study
Liu et al[14], 2022	30/30	NR	NR	Intraoperation	3D of CT	RCT
Kinoshita et al[15], 2017	12/10	NR	NR	Pre-operation	3D of CT	Cohort study
Wang[16], 2019	52/52	47.34 ± 4.08/47.15 ± 4.16	27/25, 26/26	Pre-operation and intraoperation	3D of ultrasound and CT	Cohort study
Huang et al[17], 2016	60/60	48.6 ± 5.2/47.5 ± 6.8	28/32, 31/29	Pre-operation and intraoperation	3D of ultrasound and CT	RCT
Huang et al[18], 2016	40/40	63.45 ± 3.62/62.97 ± 3.59	21/19, 22/18	Intraoperation	3D of ultrasound	RCT
Huang et al[19], 2020	30/30	51.95 ± 17.58/51.67 ± 17.35	22/8, 21/9	Intraoperation	3D of ultrasound	RCT
Chen et al[20], 2020	80/80	NR	46/34, 45/35	Pre-operation and intraoperation	3D of ultrasound and CT	RCT
Yang et al[21], 2023	20/36	66.5 ± 12.4/62.4 ± 10.9	9/11, 23/13	Pre-operation and intraoperation	3D of ultrasound and CT	Cohort study

3D: Three-dimensional; CT: Computerized tomography; NR: Not reported; MR: Magnetic resonance; RCT: Randomized controlled trial.

Table 2 Article quality assessment by Newcastle-Ottawa Scale: Cohort study.

Ref.	Representativeness of the exposed cohort (1)	Selection of the non-exposed cohort (1)	Ascertainment of exposure (1)	Demonstration that outcome of interest was not present at start of study	Compare ability of cohorts on the basis of the design or analysis (2)	Assessment of outcome (1)	Was follow up long enough for outcomes to occur (1)	Adequacy of follow up of cohorts (1)	Total
Liu et al[22], 2023	1	1	1	1	2	1	1	1	9
Zhou et al[11], 2021	1	1	1	1	2	1	1	1	9
Li et al[10], 2020	1	1	1	1	2	1	1	1	9
Zhou et al[24], 2022	1	1	1	1	1	1	1	1	8
Wang[16], 2019	1	1	1	1	2	1	1	1	9
Yang et al[21], 2023	1	1	1	1	1	1	1	1	8
Kinoshita et al[15], 2017	1	1	1	1	0	1	1	1	7

Meta-analysis results

Surgery time: A total of 12 studies were included[10-21], all of which provided detailed reports on surgery time (Figure 3A), with 478 participants in the 3D group and 495 in the traditional group. The results showed considerable heterogeneity across the studies (P < 0.00001, I² = 99%), therefore, a random effects model was applied. The meta-analysis results indicated: MD = -22.56; 95%CI: -29.30 to -15.81, P < 0.00001, the surgery time in the 3D group was significantly shorter than that in the traditional group.

Figure 3 Meta-analyses of surgical and postoperative outcomes (including operation time, puncture-related indicators, and postoperative complications). A: Meta-analysis of operation time; B: Meta-analysis of number of punctures; C: Meta-analysis of first puncture success rate; D: Meta-analysis of overall puncture success rate; E: Meta-analysis of post-operative bleeding; F: Meta-analysis of biliary fistula; G: Meta-analysis of postoperative infection; H: Meta-analysis of overall post-operative complications rate; I: Meta-analysis of the reduction levels of postoperative total bilirubin. 3D: Three-dimensional; CI: Confidence interval.

Number of punctures: A total of 10 studies were included[10-15,17,19-21], all of which reported the number of punctures in detail (Figure 3B), with 437 participants in the 3D group and 446 in the traditional group. The results showed notable heterogeneity among the studies (P < 0.00001, I² = 98%), consequently, a random effects model was utilized. The findings from the meta-analysis indicated: MD = -1.29; 95%CI: -1.83 to -0.75, P < 0.00001, the number of punctures in the 3D group was significantly lower than that in the traditional group.

First puncture success rate: A total of 6 studies were included[10,14,18,19,22,23], all of which reported in detail the first puncture success rate (Figure 3C), with 286 people in the 3D group and 282 in the traditional group. There was no considerable variation among the studies (P = 0.25, I² = 24%), a fixed-effect model was employed. The outcomes of the meta-analysis demonstrated: OR = 5.66; 95%CI: 3.27-9.80, P < 0.00001, the first puncture success rate in the 3D group was significantly higher than that in the traditional group.

Overall puncture success rate: A total of 8 studies were included[11-13,15,18,20,22,24], all of which reported detailed data on the overall puncture success rate (Figure 3D), with 445 people in the 3D group and 375 people in the traditional group. The results showed no notable heterogeneity in the studies (P = 0.68, I² = 0%), thus, a fixed-effect Model was implemented. The meta-analysis findings suggested: OR = 3.61; 95%CI: 1.98-6.55, P < 0.0001, the overall puncture success rate in the 3D group was significantly higher than that in the traditional group.

Post-operative bleeding: A total of 12 studies were included[10-21], all of which reported post-operative bleeding in detail (Figure 3E), with 466 patients in the 3D group and 483 patients in the traditional group. The studies showed no notable differences (P = 0.97, I² = 0%), a fixed-effect model was adopted. The results derived from the meta-analysis revealed: OR = 0.28; 95%CI: 0.15-0.53, P < 0.0001, post-operative bleeding occurred significantly less in the 3D group than that in the traditional group.

Biliary fistula: A total of 7 studies were included[10-12,16-18,20], all of which reported in detail on biliary fistula (Figure 3F), with 354 people in the 3D group and 357 in the traditional group. There was no considerable variation among the studies (P = 1.00, I² = 0%), hence, a fixed-effect model was implemented. According to the meta-analysis results: OR = 0.29; 95%CI: 0.12-0.69, P = 0.005, the incidence of postoperative biliary fistula was significantly lower in the 3D group than that in the traditional group.

Postoperative infection: A total of 11 studies were included[10-14,16-21], all of which provided detailed reports on postoperative infections (Figure 3G), with 466 people in the 3D group and 485 people in the traditional group. The studies showed no notable differences (P = 0.99, I² = 0%), a fixed-effect model was applied. The findings from the meta-analysis indicated: OR = 0.35; 95%CI: 0.18-0.69, P = 0.003, the incidence of postoperative infections in the 3D group was significantly lower than that in the traditional group.

Overall post-operative complications rate: A total of 14 studies were included[10-12,14-24], all of which reported the overall post-operative complications rate in detail (Figure 3H), with 710 people in the 3D group and 660 in the traditional group. There was no considerable heterogeneity across the studies (P = 0.74, I² = 0%), consequently, a fixed-effect model was implemented. The outcomes of the meta-analysis demonstrated: OR = 0.25; 95%CI: 0.17-0.36, P < 0.00001, the overall post-operative complications rate in the 3D group was significantly lower than that in the traditional group.

Bile duct drainage effect: A total of 4 studies were included[14,19,21,23], all of which reported in detail the reduction levels of postoperative total bilirubin (Figure 3I), with 143 people in the 3D group and 160 people in the traditional group. The results showed no marked heterogeneity in the studies (P = 0.44, I² = 0%), a fixed-effect model was employed. The results of the meta-analysis demonstrated: MD = -1.38; 95%CI: -3.29 to 0.53, P = 0.16, there was no significant difference in the reduction levels of postoperative total bilirubin between the 3D group and the traditional group. A total of 2 studies were included[19,23], both of which reported in detail the reduction levels of aspartate transaminase and alanine transaminase (Figure 4A and B), with 93 patients in the 3D group and 94 patients in the traditional group. The results showed no significant heterogeneity among the studies (P = 0.97, I² = 0%) and (P = 0.97, I² = 0%), consequently, a fixed-effect model was implemented. The findings from the meta-analysis indicated: MD = -0.71; 95%CI: -48.11 to 46.70, P = 0.98 and MD = 2.20; 95%CI: -44.61 to 49.01, P = 0.93, indicating that there was no significant difference in the reduction levels of aspartate transaminase and alanine transaminase between the 3D group and the traditional group.

Figure 4 Meta-analysis of the reduction levels of aspartate transaminase and alanine transaminase. A: Meta-analysis of the reduction levels of aspartate transaminase; B: Meta-analysis of the reduction levels of alanine transaminase. 3D: Three-dimensional; CI: Confidence interval.

Subgroup analysis

Subgroup analysis were conducted on the surgery time according to guidance stages of the 3D reconstruction, 3D reconstruction imaging modalities, and types of studies. The results showed no significant reduction in the heterogeneity across subgroups. The overall meta-analysis outcomes remained stable and consistent (Figure 5A-C).

Figure 5 Meta-analysis for subgroup analysis of surgery time. A: Meta-analysis for subgroup analysis of surgery time according to guidance stages of the three-dimensional reconstruction; B: Meta-analysis for subgroup analysis of surgery time according to three-dimensional reconstruction imaging modalities; C: Meta-analysis for subgroup analysis of surgery time according to types of studies. CI: Confidence interval.

Sensitivity analysis

For the high heterogeneity of the meta-analysis, sensitivity analysis was conducted by excluding one study at a time to assess the impact of study quality on the stability of the meta-analysis. The results showed that no single study had a significant impact on the research findings. This indicates that our meta-analysis results are reliable.

Publication bias

Publication bias was assessed using a funnel plot based on the overall post-operative complications rate (Figure 6). The plot showed a generally symmetrical distribution, with all studies falling within the area bordered by the two diagonal lines, suggesting no significant publication bias.

Figure 6 Funnel plot based on overall post-operative complications rate. OR: Odds ratio.

DISCUSSION

In traditional PTBD, the surgeon mainly locates the puncture point by interpreting two-dimensional imaging data, and then uses X-ray or ultrasound for guidance during the operation. The entire surgery mainly relies on the surgeon’s anatomical knowledge and subjective judgment[25-27]. However, in patients with obstructive jaundice, liver morphology and vascular anatomy often vary significantly. In such cases, blindly performing programmed puncture operations will bring unpredictable risks to the patient[28].

With the continuous development of modern information technology in the medical field, 3D reconstruction technology has gradually become an important technical means to guide PTBD. It mainly reprocesses traditional two-dimensional imaging data, using advanced segmentation algorithms to accurately extract contour information of key anatomical structures, such as the liver and bile ducts. Additional image processing techniques, including denoising and enhancement, and artifact removal, help to restore the true morphology of tissues. Through multi-modal image fusion, this process generates a detailed, individualized 3D model, offering surgeons a comprehensive view of the patient’s liver anatomy[29]. This personalized model enables surgeons to visualize lesion locations and bile duct pathways, simulate surgical procedures in advance, evaluate different puncture paths, and optimize surgical plans. Consequently, it enhances both the precision and safety of PTBD procedures[7,8].

The success of PTBD depends critically on puncture point selection and real-time needle trajectory adjustment[12,30]. 3D reconstruction offers substantial advantages in both areas. Preoperatively, it allows for a thorough understanding of biliary anatomy - including branching patterns, directions, and vascular relationships - facilitating accurate puncture planning while minimizing vascular injury[10,11,13,22]. Intraoperatively, the real-time 3D model provides dynamic navigation, enabling the surgeon to fine-tune the needle’s angle and depth to ensure successful cannulation and reduce the risks associated with repeated blind punctures[10,13]. The result of this study fully confirms this point. The application of 3D reconstruction technology significantly improves the puncture success rate of PTBD, effectively shortens the operation time and reduces the number of punctures, which has obvious advantages compared with the traditional methods. Previous studies have shown that prolonged operation time and increased number of punctures are important factors leading to a high incidence of postoperative complications[21,22,31]. The results of this study are consistent with this. The incidence of postoperative bleeding, biliary fistula and infection in the traditional group is significantly higher than that in the 3D reconstruction group. In some cases, even death occurs due to serious complications, highlighting the important value of 3D reconstruction technology in improving surgical safety[10].

Effective drainage is the top priority of PTBD, and the accurate selection of the target bile duct is essential to achieving it[32]. In theory, the high-resolution anatomical visualization provided by 3D reconstruction should enhance target duct identification, supporting better drainage outcomes[10]. Some studies have confirmed this theoretical advantage[24]. However, in this meta-analysis, 3D reconstruction technology did not show a significant advantage in the effect of bile drainage. This result may be related to the following factors. First is lacking of real-time adaptability. Preoperative 3D models do not account for dynamic anatomical changes caused by respiratory motion or patient positioning during surgery, which may reduce puncture accuracy[13]. Then, the absence of stratification could be the other reason. Most included studies did not stratify patients by bile duct dilation degree or obstruction site, potentially masking the subgroups most likely to benefit. Finally, the limited number of studies assessing bile drainage introduces risk of bias and limits statistical power. Despite this, the observed trend in favor of 3D reconstruction indicates potential clinical value. Future large-scale, multicenter randomized controlled trials - stratified by anatomical classification and obstruction grading - are needed to clarify the technology’s impact on drainage efficacy.

This study included 15 studies, 13 of which were randomized controlled trials or prospective cohorts with high methodological quality. However, significant heterogeneity was detected in several outcomes, particularly surgery time (I² = 99%). Subgroup analyses based on guidance stages of 3D reconstruction, imaging modalities, and study types did not significantly reduce heterogeneity. Nonetheless, the direction of the effect sizes across all subgroups consistently favored the 3D group, indicating robustness of the findings.

The residual heterogeneity may be attributed to uncontrolled primary heterogeneity sources, and limited subgroup sample sizes. Differences in surgeon experience, puncture techniques, disease types, study designs, and reconstruction software may have had a greater impact than the subgroup variables. Insufficient numbers of studies in each subgroup reduced the sensitivity of the I² statistic, limiting the detection of meaningful heterogeneity changes. Despite the unresolved heterogeneity, the consistency in effect direction strengthens the reliability of our overall conclusions.

In conclusion, 3D reconstruction technology offers significant benefits in PTBD for patients with obstructive jaundice, especially in improving puncture success rate and safety. Although its impact on bile drainage effectiveness has not been fully demonstrated, preliminary evidence suggests positive potential. Therefore, 3D-assisted PTBD has broad clinical application. Future studies should focus on large-scale, multicenter, and high-quality trial designs, ideally stratified by anatomical and clinical factors, to refine its application strategies and further improve the treatment outcomes for obstructive jaundice.

CONCLUSION

3D reconstruction technology significantly improves the puncture success rate and safety of PTBD. However, it has no significant advantage in bile drainage effectiveness. Continued research is warranted to further explore its clinical value and optimize its application.

ACKNOWLEDGEMENTS

We would like to thank all of the study participants.