Intraoperative “white test” for bile leak detection during liver resection: a video vignette

Abstract

Postoperative bile leak remains a significant source of morbidity after liver resection, especially following major or complex hepatectomies. The intraoperative “white test” is a simple technique to improve detection of bile leaks through retrograde injection of a lipid emulsion into the biliary tree, enabling direct visualization of open bile ducts on the transection surface. We report its use during a minimally invasive extended left hepatectomy for colorectal liver metastases. After parenchymal transection, the white test was performed via the cystic duct stump using Intralipid (Fresenius Kabi), identifying a focal bile leak near the left pedicle stapling line, which was promptly secured. The postoperative course was uneventful, with no bile leak or liver-related complications. Follow-up imaging showed no recurrence or biloma. This case highlights the technical simplicity and high visual contrast of the white test for intraoperative bile leak detection, with the accompanying video offering added educational value beyond a written description.

INTRODUCTION

Postoperative bile leak remains an important source of morbidity after liver resection, particularly after major or technically complex hepatectomy [1]. The intraoperative “white test,” first described by Nadalin et al. [2] in 2008, was developed as a simple method to enhance detection of bile leaks by retrograde injection of a lipid emulsion into the biliary tree, allowing direct visualization of open bile ducts on the transection surface. Compared with aqueous dyes, the technique offers strong contrast, does not stain the parenchyma and can be repeated after repair. We present its application during a minimally invasive extended left hepatectomy for colorectal liver metastases, highlighting its technical and educational value.

CASE

A 46-year-old man was referred to our hepatobiliary service for surgical management of synchronous liver metastases from a sigmoid adenocarcinoma. The tumor was mismatch-repair proficient, KRAS wild-type, and BRAF-negative. His oncologic course was complicated early by an acute hypersensitivity reaction during the first cycle of CAPOX (capecitabine and oxaliplatin) chemotherapy, which was resolved with antihistamines. Although he subsequently completed six cycles without recurrence, the event significantly influenced treatment decisions. Bevacizumab was not initiated due to persistent port-related healing issues. Given his KRAS wild-type status and the need for a better-tolerated regimen, he was transitioned to single-agent panitumumab, achieving an excellent radiographic response.

Preoperative imaging demonstrated bilobar disease involving segments IV, V, VII and VIII, with the dominant lesion infiltrating the middle hepatic vein and abutting the hilar plate. An R1 vascular margin was anticipated. The surgical strategy consisted of a laparoscopic extended left hepatectomy combined with ablation of the segment VII lesion due to its proximity to the right hepatic vein, in order to preserve the only outflow of the future liver remnant, estimated at 1,167 cm³ (59% of total liver volume).

Upon abdominal entry, the lesion in segment IVb/V was immediately visible (Fig. 1A); exploration excluded peritoneal involvement. The right liver was fully mobilized, and intraoperative ultrasound was used to delineate the course of the middle and right hepatic veins. The main lesion was confirmed to involve the middle hepatic vein and infiltrate the left pedicle. The porta hepatis was encircled with an umbilical tape to allow Pringle clamping. Calot triangle was dissected, and the cystic structures divided, leaving a long cystic duct stump for subsequent white testing. As the patient presented a replaced left hepatic artery, this was controlled within the ligamentum venosum; the left portal vein was consequently isolated and ligated. Indocyanine green (ICG) confirmed perfusion demarcation (Fig. 1B).

Fig. 1.

Intraoperative white test during laparoscopic extended left hepatectomy. (A) Metastatic lesion in segments IVb/V. (B) Indocyanine green fluorescence showing parenchymal demarcation after control of the left-sided vascular inflow. (C) Catheterization of the cystic duct stump. (D) Identification of a focal bile leak via retrograde lipid emulsion injection.

Parenchymal transection was performed using a combination of ultrasonic dissection, harmonic scalpel, and clips. As anticipated, tumor involvement of the hilar plate required stapling of the left pedicle flush at its convergence. During cranial progression of the transection toward the hepatocaval confluence, the middle hepatic vein was encountered at the junction of tributaries from segments IVb and V, together with a branch from segment VIII. In this area, clip control proved insufficient and a focal venotomy occurred, repaired with a polypropylene suture. The distal segment of the middle hepatic vein was free of disease; this was subsequently stapled before the confluence of the sectoral branches from segments VII and VIII, preserving venous drainage of the right superior liver. The left hepatic vein was then divided, completing the resection. The transection phase involved six Pringle cycles of 13–15 minutes each, separated by strict minimum 5-minute reperfusion intervals, totaling 83 minutes of liver ischemia.

On inspection of the transection surface, no bile leak was identified. A white test was therefore performed. A 4.5-French × 18-inch cholangiogram catheter was inserted into the cystic duct stump (Fig. 1C) and held in position with a large titanium clip tightened sufficiently to prevent dislodgement. Intraoperative cholangiography was not performed in this case, though it can be executed via the same catheter if anatomy requires clarification. The Pringle maneuver was reapplied, and 20 mL of Intralipid (Fresenius Kabi) 20% were manually injected retrogradely using a 50 mL syringe: this immediately demonstrated a focal bile leak adjacent to the left pedicle staple line (Fig. 1D), which was secured with a clip. Repeat testing confirmed complete sealing. Both tests contributed less than 2 minutes of additional inflow occlusion time.

Given the extent of resection, the segment VII lesion was treated with ultrasound-guided microwave ablation to preserve the right hepatic vein, without evidence of venous compromise. The specimen was extracted through a Pfannenstiel incision. Hemostasis of the transection surface was secured with topical agents, and a drain was left in situ.

The postoperative course was uneventful, with no bile leak or liver-specific complications. Follow-up laboratory values demonstrated an expected, transient postoperative transaminitis with rapid functional recovery, with all tests peaking on postoperative day 1 (alanine aminotransferase [ALT], 265 U/L; aspartate aminotransferase [AST], 504 U/L; total bilirubin, 46 µmol/L; international normalized ratio [INR], 1.2) and almost normalized by day 5 (ALT, 58 U/L; AST, 40 U/L; total bilirubin, 18 µmol/L; INR, 1.1). Pathology confirmed two colorectal liver metastases with approximately 50% treatment-related necrosis and a focal microscopic R1 vascular margin at the hilar plate, consistent with the operative strategy. Panitumumab was resumed 6 weeks after surgery, followed by chemoradiation for the primary tumor. Follow-up imaging at 1 month demonstrated no recurrence and an expected ablation zone in segment VII, with stability confirmed on positron emission tomography imaging at 3 months.

DISCUSSION

This case highlights the utility of the white test as a simple adjunct for intraoperative bile leak detection in minimally invasive liver surgery. Its principal advantage lies in the identification of small, otherwise occult biliary leakages at the completion of parenchymal transection, allowing immediate repair.

Compared with similar techniques, the white test offers several practical advantages. Methylene blue can obscure the operative field through diffuse parenchymal staining, limiting the ability to reassess after repair [3]. ICG is now widely adopted but may lack the spatial resolution required to localize small leaks on the transection surface, resulting in failure to identify the source of leakage in up to one in six cases [4]. When administered systemically, ICG has the unique advantage of detecting leakage from “orphan” ducts, disconnected from the main biliary tree [5]; however, this approach is also limited by a low signal-to-background ratio, which can make it difficult to distinguish fluorescent bile against a fluorescent liver surface [6]. This limitation can be partially mitigated by optimizing the timing of ICG administration; however, this introduces logistical challenges, as injection must precede parenchymal transection by a sufficient interval to allow hepatocellular uptake and biliary excretion. This interval may vary depending on liver function, inter-individual variability and institutional protocols, and can influence the intensity and distribution of biliary fluorescence [7]. In contrast, a lipid emulsion provides a high-contrast, localized efflux that is readily identifiable, does not stain the liver parenchyma, and can be repeated as needed.

The physical properties of lipid emulsions are particularly suited to this application. Their opacity and higher viscosity promote retention at the ductal opening, facilitating precise identification and targeted control. Unlike propofol-based techniques [8], which have been described for similar purposes, standard lipid emulsions such as Intralipid are not subject to anesthetic drug regulations, simplifying their use in routine surgical practice. Utilizing a large 50 mL Luer lock syringe for manual injection introduces an inherent mechanical safeguard. The large barrel diameter reduces the pressure transmitted by the manual plunge, which provides immediate tactile force feedback to the operating surgeon. If the biliary network is entirely sealed, the system pressure rises rapidly, creating strong physical resistance against the barrel and indicating to stop the injection. This reduces the risk of overdistension and protects the microbiliary network from disruption, a risk that increases significantly with smaller syringe volumes. Conversely, if an open leak is present on the transection surface, the defect acts as a low-pressure vent, allowing the lipid emulsion to escape and become visible without requiring high injection force.

Regarding the additional Pringle time required for the test, it must be noted that our patient was relatively young with healthy, non-cirrhotic liver parenchyma. In patients with underlying cirrhosis, severe steatosis, chemotherapy-associated liver injury, or a marginal future liver remnant, any additional ischemic burden warrants consideration. In the present case, the white test required only a brief additional period of inflow occlusion following completion of parenchymal transection. While the test can theoretically be performed without inflow occlusion by relying on the sphincter of Oddi, the lack of proximal control risks downstream escape of the lipid emulsion into the duodenum, potentially yielding a false-negative result. Alternatively, selective distal common bile duct clamping avoids global ischemia but introduces risks of iatrogenic ductal trauma, bleeding, or focal devascularization. In this setting, a short, controlled Pringle maneuver provided reliable conditions for leak detection. The decision to perform the test, and the method used to obtain biliary pressurization, should remain individualized according to liver quality, extent of resection, and overall operative context.

While a recent multicentric study did not demonstrate a statistically significant reduction in bile leak rates with routine use of this technique, interpretation is limited by a smaller-than-planned sample size, as well as heterogeneity of patient populations and variability in surgical practice [9]. Its clinical relevance is currently being further evaluated in a multicenter randomized trial in Switzerland (BiLe-Trial) [10].

In the present case, the white test enabled identification and immediate control of an otherwise occult bile leak, underscoring its potential value as a targeted intraoperative adjunct. The accompanying video (Supplementary Video 1) demonstrates the characteristic appearance and behavior of the lipid emulsion at the leak site, providing additional educational value for surgeons seeking to adopt the technique.

Supplementary materials

Supplementary materials can be found via https://doi.org/10.7602/jmis.2026.29.3.167.