Patients and Methods
Donor and Recipient Characteristics
From October 2018 to April 2019, 164 pediatric liver transplants were performed at the Organ Transplantation Center, Tianjin First Central Hospital. During the same period, 16 pediatric liver transplants using split-graft pediatric donors were performed. The liver grafts were split from 8 donation after brain death (DBD) donors younger than 7 years. The criteria and practical guidance for the determination of brain death in children formulated by the Brain Injury Evaluation Quality Control Center of the National Health and Family Planning Commission in China were used to determine brain death. The allocation of liver grafts was regulated by China Liver Transplant Registry (CLTR); the liver grafts of donors younger than 12 years were first allocated to patients younger than 12 years. Well-selected pediatric patients were matched with the donors; the donor and recipient characteristics are listed in Table 1 and Table 2. All the recipient procedures were performed at the Tianjin First Central Hospital. This study was approved by the institutional review board (approval No. 2018N135KY).
Criteria for "Split-able" Liver Grafts
Pediatric donors who met the following criteria are considered to have "split-able" livers: (1) DBD donor, (2) intensive care unit (ICU) stay <5 days, (3) no or single low-dose vasopressor therapy, (4) estimated cold ischemia time <10 hours, (5) ideal donor-to-recipient liver volume matching, and (6) serum transaminase levels <3 times normal value. If only 1 of the criteria is missing and the others are well controlled, the graft can be accepted for split.
Graft Volume Estimation
The first evaluation of the graft volume was made as soon as we received information on the precise weight of the liver graft immediately after procurement. According to previously accumulated data from the abdominal computed tomography scans of potential pediatric donors before procurement in our hospital in addition to our clinical experience, the left lateral segment in children accounts for approximately 25% to 30% of the whole liver volume (unpublished data). The density of the liver is considered to be 1.051 ± 0.013 g/mL. Therefore, we could estimate the weight of the left lateral segment (LLS) and extended right lobes (ERLs) and select suitable recipients. Once the graft arrived at our center, it was evaluated further for the possibility of split. The LLS tended to be allocated to patients weighing <7 kg. One of the liver grafts in this study was procured in our hospital and was split in situ. An abdominal computed tomography scan was performed on this patient before procurement, so the liver volume could be estimated before procurement.
Donor and Recipient Matching
One of the most important indicators for donor-to-recipient matching was graft-to-recipient weight ratio (GRWR). The ideal GRWR was 2% to 4%, and the minimal accepted ratio was 1.5%.[10,11] In some patients, a GRWR >4% was also acceptable if the abdomen could be closed without tension and postoperative ultrasound showed normal blood flow. In addition, other factors, including severity of the primary disease, degree of the portal hypertension, and patient status before transplant, were taken into consideration in deciding the appropriate recipients.
One of the splitting procedures was performed in situ (donor 4) and the others were performed ex situ (Figure 1A-D). All the split procedures were completed within 1 hour. Biopsy specimens were obtained from all the grafts. The liver grafts were divided into the LLS (segments II and III) and ERLs (segments I and IV through VIII). The LLS contained the left hepatic duct, and the ERLs included the common bile duct. The anatomy of biliary system was detected with either cholangiography or biliary tract probe before splitting. Reconstruction of the hepatic outflow, portal vein, and hepatic artery in each recipient was summarized in Table S1 (Figure 1E,F). The "piggyback" technique, like living donor liver transplant, was performed in 14 recipients, and the classic technique with inferior hepatic vena cava replacement was performed in 2 recipients (recipient 2 and 16). Duct-to-duct biliary reconstruction was performed in 1 recipient (recipient 2) who was diagnosed with progressive familial intrahepatic cholestasis. The other patients diagnosed with biliary atresia (BA) underwent roux-en-Y hepaticojejunostomy.
Split liver transplant using pediatric grafts. (A) In situ split procedure: (a) extended right lobe (ERL), (b) left lateral segment (LLS). (B) Dissection of vessels before LLS procurement: (a) ERL, (b) LLS, (c) left portal branch, (d) left hepatic artery. (C) Ex situ split procedure: (a) ERL, (b) LLS. (D) Division of left portal branch during ex situ split: (a) ERL, (b) LLS, (c) split site, (d) portal trunk, (e) left portal branch. (E) Vessel reconstruction of LLS: (a) LLS, (b) portal vein, (c) hepatic artery. (F) Vessel reconstruction of ERL: (a) ERL, (b) portal vein, (c) hepatic artery
The immunosuppressive protocol consisted of tacrolimus and steroids. Tacrolimus was started on the second or third postoperative day. The target blood trough level of tacrolimus was 7 to 10 ng/mL during the first month, 5 to 8 ng/mL for the second through sixth months, and 5 ng/mL thereafter. Steroids were given at the time of graft implant with a dose of 10 mg/kg, then gradually reduced, and completely weaned by 6 months after transplant as long as liver function was stable.
Blood tests and abdominal ultrasound were performed every day in the first postoperative week to monitor functional recovery and blood flow of the liver grafts. Liver biopsy was conducted in patients whose serum transaminase levels were >2 times normal value. After discharge, the patients underwent routine tests every week in the first month and every 2 weeks from the second through third months. Results were collected and analyzed by our specialized follow-up team.
Continuous data are presented as mean/standard deviation or median/range and compared with the use of paired t test or Mann-Whitney U test according to the distribution. Categorical data are presented as number and ratio and compared with the use of a χ 2 test. SPSS 23.0 software (IBM, München, Germany) was used to process the data. The value of P < .05 was considered statistically significant.
American Journal of Transplantation. 2020;20(3):817-824. © 2020 Blackwell Publishing