Systemic indocyanine green administration to detect bile leakage after liver surgery: a prospective clinical trial, using historical controls "A Novel Approach for Minimizing Postoperative Biliary Complications"

ISM2025-04
11 June,2025

We are pleased to announce that a research group led by Senior Lecturer Takehiko Hanaki (currently affiliated with the Department of Medical Education), Professor Yoshiyuki Fujiwara, both from the Department of Gastrointestinal and Pediatric Surgery, Faculty of Medicine, Tottori University (Yonago, Tottori, Japan), and Professor Hisashi Noma from the Institute of Statistical Mathematics (Tachikawa, Tokyo, Japan), has conducted a prospective clinical trial to evaluate the effectiveness of a novel intraoperative assessment method using indocyanine green (ICG) for detecting and repairing bile leakage during liver resection surgery. 

This method utilizes systemically administered ICG in conjunction with a near-infrared (NIR) camera. The study demonstrated that this approach enables high-sensitivity intraoperative detection and repair of minute bile leaks—leaks that would typically go unnoticed by the naked eye. Furthermore, the method significantly reduced the incidence of postoperative bile leakage, providing clear evidence of its clinical benefit. 

To our knowledge, this is the first study in the world to clinically demonstrate that detecting leakage of systemically administered ICG can lead to a reduction in postoperative bile leakage after liver resection surgery.

Key Findings of This Study

• This study is the first in the world to clinically demonstrate the effectiveness of a novel intraoperative technique that uses systemically administered indocyanine green (ICG) and near-infrared (NIR) imaging to visualize and repair bile leakage during liver surgery.
• A prospective clinical trial was conducted, comparing patients who received ICG-guided treatment with a historical control group who underwent standard care. The effectiveness of the intervention was evaluated using statistical causal inference methods.
• In the ICG group, bilirubin concentration in drain fluid on postoperative day 3 was significantly lower, and the incidence of bile leakage was markedly reduced.
• ICG fluorescence imaging detected 11 times more bile leak sites compared to visual inspection with the naked eye, allowing for earlier and more precise repairs.
• No adverse effects associated with ICG administration were observed. The method was confirmed to be highly safe, contributing to shorter hospital stays and fewer postoperative complications.

Background of the Study
 Liver resection is widely performed as a curative treatment for liver cancer and metastatic liver tumors. However, bile leakage, a postoperative complication, remains a significant clinical challenge (Figure 1).

 Bile leakage can lead to serious complications such as intra-abdominal infection, bile peritonitis, and liver failure, potentially requiring extended hospitalization, reoperation, or secondary interventions such as re-drainage. 

 Traditionally, intraoperative detection of bile leakage has relied on visual inspection, which often fails to identify small or subtle leaks. In addition, conventional leak tests that involve injecting contrast agents or dyes into the biliary tract are invasive and carry risks of bile duct injury and infection. 

 Against this background, a novel method has garnered attention in recent years: systemic administration of indocyanine green (ICG), which is naturally excreted into bile from the liver, allowing non-invasive visualization of bile leakage using a near-infrared fluorescence camera. However, previous studies have been limited to single-center retrospective analyses or case reports, and there has been no prospective clinical trial-based evidence to validate its clinical utility. 

 This study represents the first prospective clinical trial to evaluate the effectiveness of this technique—detecting and repairing bile leakage intraoperatively using systemically administered ICG—with the aim of establishing a new standard in the prevention of postoperative bile leakage.

Overview of the Research Findings
 This study was a prospective clinical trial aimed at preventing postoperative complications by enabling the early detection and intraoperative repair of bile leakage following hepatectomy.
 The research team investigated a technique in which the fluorescent dye indocyanine green (ICG) was administered systemically. By taking advantage of ICG's natural excretion into bile, the team was able to visualize bile leakage from the liver resection surface using near-infrared (NIR) imaging (Figure 2).

 Specifically, patients undergoing hepatectomy received an intravenous injection of 10 mg of indocyanine green (ICG) before liver parenchymal transection. The resection surface was then examined intraoperatively using fluorescence imaging to identify sites of bile leakage, which were immediately treated by suturing or other appropriate measures. This method enabled the high-sensitivity detection of small bile leaks that were otherwise undetectable by conventional visual inspection, thereby facilitating real-time intraoperative management. 

 In this trial, the research team compared a prospective intervention group (n = 40), in which ICG was used, with a historical control group (n = 44) who underwent standard surgery without ICG. The primary endpoint was the total bilirubin concentration in drain fluid on postoperative day 3, a widely accepted surrogate marker for bile leakage as defined by the International Study Group of Liver Surgery (ISGLS). 

 For statistical analysis, to reduce bias inherent in non-randomized studies and accurately estimate treatment effects, the study employed inverse probability of treatment weighting (IPTW) based on propensity scores, within the framework of statistical causal inference. 

 As a result, the ICG group showed a significant reduction in total bilirubin concentration in drain fluid (−1.11 mg/dL), and the incidence of bile leakage of Grade A or higher was markedly lower (5.0% in the ICG group vs. 27.3% in the control group).
 Notably, no clinically significant bile leaks (Grade B or higher) occurred in the ICG group. 

 Furthermore, the number of bile leak sites detected using ICG fluorescence imaging was approximately 11 times higher than that identified by visual inspection alone. Notably, this method allowed for the detection of Type D bile leaks according to the Nagano classification, which originate from bile ducts not in continuity with the common bile duct—leaks that are often missed using conventional techniques.

 ICG is a well-established and highly safe dye traditionally used in liver function testing, and no adverse effects were reported in this study. In addition, the ICG group demonstrated a shorter average hospital stay (6.1 days vs. 17.6 days), and a reduction in postoperative complications of Grade II or higher according to the Clavien–Dindo classification. 

 These findings suggest that ICG-guided intraoperative assessment enables early detection and prevention of bile leakage, a major postoperative complication, thereby potentially promoting faster recovery and more efficient use of medical resources. 

 This study provides the first clinical evidence in the world supporting the use of ICG as a new intraoperative standard in liver surgery.

Study Details
 This study evaluated the effectiveness and safety of a novel intraoperative assessment technique for the early detection and repair of bile leakage following liver resection. The method combines systemically administered indocyanine green (ICG) with near-infrared (NIR) fluorescence imaging. 

 Conducted as a prospective, single-center clinical trial at Tottori University Hospital, the study compared an intervention group (n = 40), in which ICG was used intraoperatively, with a historical control group (n = 44) that received standard care without ICG guidance. 

 In the intervention group, 10 mg of ICG was administered intravenously during surgery, and its excretion into bile allowed for real-time visualization of the liver resection surface using a NIR camera (Figure 3). This enabled precise identification of bile leakage sites, which were then immediately sutured or otherwise repaired intraoperatively.

 The primary endpoint of this study was the total bilirubin concentration in drain fluid on postoperative day 3 (POD 3), which is an internationally recognized diagnostic indicator for bile leakage.
 To minimize potential bias due to differences in surgical techniques or patient background factors, the analysis employed inverse probability of treatment weighting (IPTW) based on propensity scores.

 As a result, the intervention group showed:

• A significant reduction in total bilirubin concentration in drain fluid (−1.105 mg/dL, P < 0.001).
• A reduced incidence of bile leakage of Grade A or higher, at 5.0% (compared to 27.3% in the control group, P = 0.002).
• No cases of clinically significant bile leakage (Grade B or higher),
• Approximately 11 times more leak sites detected with ICG fluorescence imaging than with visual inspection alone.
• A significantly shorter hospital stay (6.1 days vs. 17.6 days, P = 0.020).
• No adverse events or side effects related to ICG administration.

 Previously, Hanaki et al. reported the potential of ICG fluorescence imaging to sensitively detect bile leakage that is invisible to the naked eye, thereby contributing to its prevention (doi.org/10.1002/ccr3.5942, video link: https://x.gd/f3Q05). In that study, systemic ICG administration enabled early intraoperative detection and repair of micro bile leaks at the liver resection surface. The present study builds upon that foundation and is the first in the world to evaluate this approach prospectively and systematically.

 Since bile leakage is a relatively infrequent postoperative complication, statistically validating the preventive effect through a randomized controlled trial (RCT) would require a very large sample size (several hundred cases or more), which presents practical challenges.

 This study therefore holds significant clinical and methodological value, as it addresses a difficult-to-evaluate clinical issue through the use of historical controls and advanced data science techniques within a statistical causal inference framework, providing new evidence that would otherwise be hard to obtain via conventional RCTs.

Researcher’s Comment
Name: Takehiko Hanaki
Affiliation: Dept. of Medical Education / Dept. of Gastrointestinal and Pediatric Surgery, Faculty of Medicine, Tottori University
Position: Senior Lecturer
Degree: Ph.D. (Medicine)
Comment:In this study, we demonstrated that systemic administration of indocyanine green (ICG) followed by intraoperative fluorescence imaging enables the high-sensitivity detection of minute bile leaks that are often missed by conventional visual inspection. More importantly, this technique allows for real-time intraoperative repair, contributing to the prevention of postoperative bile leakage and associated complications, as well as a potential reduction in hospital stay and patient burden. 

 Our group has previously reported the utility of ICG fluorescence in surgical guidance, and through this clinical trial, we were able to scientifically and statistically validate those earlier findings. Presenting a minimally invasive and safe technique to address a serious complication such as bile leakage is a clinically meaningful advancement in improving the quality of liver surgery.
Since this method relies on equipment and agents already in clinical use, we hope it will contribute to safer and more reliable hepatectomy procedures for patients in the near future.

Publication Information
Journal: BMJ Open
Title: Systemic indocyanine green administration to detect bile leakage after liver surgery: A prospective clinical trial, using historical controls
Japanese Title:「肝切除後の胆汁漏検出における全身投与ICGの有用性：ヒストリカルコントロールを用いた前向き臨床試験」
DOI: https://doi.org/10.1136/bmjopen-2024-097205
Authors: Takehiko Hanaki¹², Keisuke Goto¹, Naruo Tokuyasu¹, Yusuke Endo³, Hiroshi Sunada³, Hisashi Noma⁴, Mikiya Kishino¹, Takuki Yagyu¹, Ei Uchinaka¹, Yuki Murakami¹, Kozo Miyatani¹, Kyoichi Kihara¹, Tomoyuki Matsunaga¹, Manabu Yamamoto¹, Teruhisa Sakamoto¹, Toshimichi Hasegawa¹, Yoshiyuki Fujiwara¹
Corresponding Author: Takehiko Hanaki (Principal Investigator)
Affiliations:
1. Department of Gastrointestinal and Pediatric Surgery, Tottori University Faculty of Medicine, Yonago, Tottori, Japan
2. Department of Medical Education, Tottori University Faculty of Medicine, Yonago, Tottori, Japan
3. Department of Advanced Medicine, Innovation, and Clinical Research Centre, Tottori University Hospital, Yonago, Tottori, Japan
4. Department of Interdisciplinary Statistical Mathematics, The Institute of Statistical Mathematics, Tachikawa, Tokyo, Japan