Routine White Test for Early Detection of Bile Leak in Elective Liver Resection (DELiveR Trial)

Bile leakage remains a clinically important complication following liver resection and is associated with increased postoperative morbidity, need for radiological, endoscopic or surgical intervention, prolonged hospitalisation and increased healthcare costs. Effective intraoperative identification and closure of open bile ducts is therefore an important component of strategies intended to reduce postoperative bile leakage.

Conventional intraoperative methods, such as visual inspection of the liver transection surface and application of sterile white gauze, rely on passive identification of visible biliary outflow and may fail to identify small or occult leakage sites. The White Test is an active intraoperative bile leak assessment involving retrograde injection of a white lipid emulsion into the biliary tree. The resulting transient increase in intrabiliary pressure may enable visualisation of white emulsion at open bile ducts on the liver transection surface and permit immediate clipping, ligation or suturing.

Previous comparative studies and meta-analyses have suggested that the White Test may reduce postoperative bile leakage. However, the available evidence is heterogeneous, and the effectiveness and safety of routine White Test use across both major and minor liver resections have not been conclusively established.

The DELiveR Trial is a single-centre, prospective, randomised, parallel-group, participant- and outcome-assessor-blinded superiority trial. The trial will include adults undergoing elective major or minor liver resection, including anatomical and non-anatomical procedures.

Written informed consent will be obtained before surgery. Final eligibility will be confirmed intraoperatively after confirmation of resectability, completion of liver parenchymal transection and achievement of haemostasis, and confirmation that a cystic duct or cystic duct stump can be safely approached for attempted cannulation. Patients found to have unresectable disease, to require a procedure incompatible with the trial protocol or to have no safely accessible cystic duct or cystic duct stump will not undergo randomisation and will be recorded as screening failures.

Eligible participants will be randomised in a 1:1 ratio to the White Test or standardised white gauze assessment using a computer-generated sequence with permuted blocks. Allocation concealment will be maintained using sequentially numbered, opaque, sealed envelopes. The envelope will be opened only after final intraoperative eligibility has been documented and immediately before the assigned intraoperative bile leak assessment.

Participants and the investigator assessing the primary outcome will remain blinded to treatment allocation. The operating surgical team cannot be blinded after allocation because of the nature of the intervention. However, allocation will not be revealed until liver parenchymal transection and haemostasis have been completed. The blinded outcome assessor will review a predefined adjudication dataset from which treatment allocation and other allocation-revealing information will be withheld.

In the intervention group, the cystic duct or cystic duct stump will be cannulated after completion of liver parenchymal transection and haemostasis. The distal extrahepatic bile duct will be temporarily occluded below the cystic duct insertion. SMOFlipid 5% will then be administered retrogradely in a volume of 10-20 mL using slow, manual, low-pressure injection. Forceful injection will not be permitted, and injection will be immediately discontinued if abnormal resistance is encountered. The liver transection surface will be inspected for leakage of white emulsion. Identified leakage sites will be clipped, ligated or sutured, and the White Test may be repeated until no further leakage is observed. After completion of the test, the biliary tree will be flushed with 20-50 mL of sterile 0.9% sodium chloride solution.

If a participant is randomised to the White Test group but cannulation or completion of the test subsequently proves technically impossible, the participant will remain in the intervention group for the primary intention-to-treat analysis. The reason for technical non-completion and the procedure actually performed will be prospectively recorded.

In the control group, sterile white surgical gauze will be systematically applied to the liver transection surface after completion of liver parenchymal transection and haemostasis. The gauze and the transection surface will be inspected for visible biliary staining. Identified leakage sites will be clipped, ligated or sutured, and the assessment may be repeated until no further visible leakage is identified. The control assessment will be conducted according to a written study-specific standard operating procedure. Alternative active intraoperative bile leak tests will not be permitted as part of the study procedure in the control group. Any clinically necessary deviation will be prospectively recorded.

All participants will have an intra-abdominal drain placed at the liver resection site. Serum and drain-fluid total bilirubin concentrations will be measured concurrently on postoperative day 3 and may be repeated thereafter while the drain remains in place or when bile leakage is clinically suspected. Serum amylase and lipase concentrations will be measured on postoperative day 1. Postoperative pancreatitis, bile duct injury, cholangitis, allergic reactions and other potentially intervention-related adverse events will be prospectively recorded.

The primary outcome is postoperative bile leakage within 30 days, defined according to the International Study Group of Liver Surgery criteria as drain-fluid bilirubin concentration at least three times the concurrent serum bilirubin concentration on or after postoperative day 3, or the need for an intervention due to a biliary collection or biliary peritonitis.

Outcome assessment will be based on postoperative biochemical measurements, the clinical course during the index hospitalisation, imaging findings, postoperative interventions, readmissions and available medical documentation during the first 30 postoperative days. Telephone follow-up after discharge will be used to identify interval events but will not be used as the sole basis for excluding postoperative bile leakage.

The primary analysis will be conducted according to the intention-to-treat principle in the overall randomised population. Logistic regression will be used with treatment allocation as the independent variable. The treatment effect will be reported as an odds ratio with a 95% confidence interval and a two-sided p value. Event rates and the absolute risk difference will also be reported.

Major versus minor hepatectomy will be evaluated only in prespecified exploratory subgroup analyses. Potential heterogeneity of the treatment effect will be evaluated using a treatment-by-hepatectomy-type interaction term. A supplementary per-protocol analysis will include participants who received the allocated bile leak assessment according to the protocol.

The sample size calculation assumes postoperative bile leakage rates of 15% in the control group and 4.95% in the White Test group. With a two-sided significance level of 0.05, 80% statistical power, 1:1 allocation and a superiority design, 262 randomised participants are required. To allow for approximately 3% of participants with unavailable primary outcome data, the target sample size is 270 participants.

No formal interim efficacy analysis or formal statistical stopping boundary is planned. An independent Data and Safety Monitoring Committee will review accumulated safety data after every 50 randomised participants or at least every six months. The Committee may recommend continuation, modification, temporary suspension or termination of the trial if an unexpected safety signal is identified.