Data collection

The study had a single-center retrospective design and was approved by the Institutional Review Board of the Affiliated Hospital of Qingdao University (IRB approval number: QYFYWZLL 28150). All grafts were sourced from brain-dead donors. Before organ donation, consent was obtained from the donor or their family. Details of the examinations required were explained to the families of all donors, and written informed consent was obtained in accordance with the principles of the Declaration of Helsinki (Edinburgh Revision, 2000).

Clinical and imaging data collected between September 1, 2022 and July 31, 2023 for patients who received liver transplantation were reviewed to identify those who met our study eligibility criteria. The inclusion criteria were as follows: ultrasound and CEUS examinations completed for the brain-dead donor on the same day as surgery to harvest the liver; recipient underwent whole liver transplantation at our hospital; and CEUS image stacks recorded in DICOM format available. The following exclusion criteria were applied: age younger than 18 years; corresponding recipient underwent split liver or multiple organ transplantation; CEUS image quality too poor for analysis. The study enrollment process is summarized in Figure 1 .

Flow chart showing the patient enrollment process. CEUS, contrast-enhanced ultrasound; EAD, early allograft dysfunction; LT, liver transplantation.

Information was collected on the following variables previously mentioned in the literature to be potential risk factors for EAD: donor characteristics (including age, sex, blood type, length of stay in the ICU), 24-hour preoperative laboratory indicators (including γ-glutamyl transpeptidase, serum creatinine, and hemoglobin), intraoperative data (including operation time, anhepatic phase, and cold ischemia time), and recipient characteristics (including age, sex, and blood type). Noteworthy is that our center obtains donor liver tissue for biopsy during the operation. And the degree of steatosis was graded and recorded based on the pathology results.

The study participants were divided into those who developed EAD (the EAD group) and those who did not (the non-EAD group). EAD was defined as the presence of one or more of the following [ ]: bilirubin >10 mg/dL on day 7 after liver transplantation; an international normalized ratio >1.6 on day 7 after liver transplantation; and an aspartate aminotransferase (AST) or alanine aminotransferase level >2000 U/L within the first 7 days after liver transplantation.

US and CEUS examinations

Ultrasound and CEUS were performed by a trained radiologist with 6 years of experience who was unaware of the clinical information and test results for the donors. All donors underwent ultrasound and CEUS examinations while maintaining stable vital signs after brain death. The study data were obtained using an iU Elite Ultrasound system (Philips Healthcare, Eindhoven, Netherlands) with a C5-1 transducer. The ultrasound scanning protocol consisted of both grayscale ultrasound and Doppler ultrasound. Grayscale ultrasound was performed to observe liver size, echotexture of the hepatic parenchyma, and presence of tumors in the liver. Doppler ultrasound was performed to explore vascularity in vascular structures. During the ultrasound examinations, hepatic artery resistive index, systolic acceleration time in the hepatic artery, portal venous flow, and portal venous velocity were measured and recorded.

In the subsequent CEUS examination, the same examiner performed a longitudinal scan of the liver parenchyma via a right intercostal approach, clearly identifying the main portal vein, right branch, and liver parenchyma. The probe was fixed in the long-axis section of the right branch of the portal vein, after which the examiner switched to the dual-phase contrast imaging mode and used a low mechanical index (0.06). A bolus injection of a second-generation contrast agent (SonoVue; Bracco, Milan, Italy) diluted in 2.0 mL of saline was administered via a peripheral vein, followed by 5.0 mL of saline solution. The timer was started at the same time as injection of the contrast agent, and dynamic perfusion of the liver was recorded for at least 2 minutes. Injection of the contrast agent was repeated, the entire liver was scanned dynamically, and any areas of abnormal enhancement were observed.

Imaging analysis

Dynamic imaging data were extracted in DICOM format using QLAB software (Philips Healthcare). One of the radiologists, who was blinded to all clinical information, independently analyzed the images and video recordings. The regions of interest (ROIs) in the liver parenchyma were 3 cm below the liver capsule and major intrahepatic vessels; they were located at a similar depth to minimize the effect of ultrasound beam attenuation. We obtained a time–intensity curve in the arterial phase for 2 minutes at each ROI and measured the ascending slope, time to peak, peak intensity (PKI), area under the curve (AUC), mean transition time, and rising time. The mean values of the parameters in the three ROIs were calculated for statistical analysis. Owing to the lack of inconsistency in the peripheral vein selected for injection of contrast agent, the interval between the first arrival time in the hepatic artery and the first arrival time in the portal vein was selected as the reference index, which we defined as the hepatic arterial–portal arrival interval (APAI) as follows: first arrival time of portal vein − first arrival time of hepatic artery [ ] ( Fig. 2 ).

Findings on contrast-enhanced ultrasound examination in a donor liver, the recipient of which developed early allograft dysfunction in the early postoperative period. The donor was a 62-year-old man with intracerebral hemorrhage and the recipient was a 49-year-old man with secondary hepatic malignancy. (a) Arrival of contrast medium in the hepatic artery at 21 seconds (white arrows). (b) Arrival of contrast medium in the portal vein at 34 seconds (white arrows). In this case, the hepatic arterial–portal arrival interval was calculated to be 13 seconds. The blue arrow indicates the regions of interest in the liver parenchyma.

Statistical analysis

Continuous variables are summarized as the mean ± standard deviation or as the median (interquartile range). Categorical variables are shown as the count (percentage). Differences in quantitative variables were compared between groups using the t -test and the Wilcoxon test. Between-group differences in qualitative variables were compared using the chi-squared test or Fisher’s exact test. Risk factors were examined in univariate analysis, and variables with a p value <0.05 were entered into a multivariate logistic regression model. The best cut-off value was determined by receiver-operating characteristic (ROC) curve analysis with calculation of sensitivity and specificity. The statistical analysis was performed using SPSS software (version 27.0; IBM Corp., Armonk, NY, USA). All tests were two-sided, and a p value <0.05 was considered statistically significant.