Portal vein interposition in pediatric living donor liver transplantation: Which vein graft type is best?

Abstract

Pediatric living donor liver transplantation (LDLT) is a technically complex operation, largely due to the small caliber and length of the vessels. Historically, all pediatric LDLT portal vein (PV) reconstructions were performed with a venous interposition, but the high rates of PV complications with cryopreserved vein conduits combined with technical modifications eliminated this need and led to end-to-end anastomoses being favored.1 PV hypoplasia is relatively common in children with biliary atresia, both due to the natural progression of the disease as well as the impact of Kasai portoenterostomy.2 This adds to the challenge of PV reconstruction in biliary atresia with interposition or jump grafts sometimes required, though portoplasty or end-to-end anastomosis to the confluence of the splenic vein and superior mesenteric vein are also options.2,3 In the article by Zamora et al,4 the authors collaborated through the Vanguard Multi-Center Study of the International LDLT group to collect data on 85 pediatric LDLTs performed for children younger than 3 years from 5 different centers in Japan, China, and Mexico. All the transplants utilized an interposition graft for the PV, and the authors compared the outcomes of different conduit options—deceased donor cold-stored iliac vein allografts versus fresh veins from the living donor or recipient. The results demonstrated that cold-stored venous allografts had more PV complications. We commend the authors for this excellent paper, which highlights the importance of the International LDLT group to investigate technical topics within LDLT that are too nuanced for most existing large national databases. At our center, we perform PV reconstructions for pediatric LDLT preferentially primarily and only use interposition conduits on rare occasions when this is not an option or if the PV anastomosis has to be revised due to an intraoperative PV thrombosis. When a conduit is necessary, we almost exclusively use deceased donor cold-stored allografts, and we therefore read this article with great interest as it would seem this is the least ideal choice. Cold-stored vessels are an appealing choice because they are readily available at centers that perform a high volume of deceased donor liver transplants. Of note, different institutions vary on the allotted time they will store deceased donor vessels in the refrigerators, adding variability to the quality of the vessels. When a conduit is needed because of an intraoperative issue, time is of the essence. Harvesting an autologous venous graft from the recipient in this circumstance can be time-consuming. Three of the 6 autologous grafts in the series were left renal veins, which is an appealing option considering that left renal vein ligation is a useful maneuver to augment PV flow when there is portal steal due to spontaneous splenorenal shunts.5 The remaining 3 autologous grafts were internal jugular vein, which adds the morbidity of a separate incision and eliminates it as an option for a future meso-Rex bypass as a rescue for PV complication.6 Obtaining extra vessels from living donors raises ethical and logistical issues. Taking the saphenous vein from a living donor adds morbidity to the operation. More commonly in this series, the gonadal or inferior mesenteric vein was used. At our center and select centers across the world, living donor hepatectomies are now being performed robotically, and additional information would be needed on the feasibility, safety, and outcomes of robotically harvested donor gonadal or inferior mesenteric vein grafts. There are several limitations to this study that make us hesitant to abandon the use of cold-stored venous allografts. First, it is not clear how many cases each center contributed. The authors state that the group gathers data from both high-volume and low-volume centers. Center volume certainly influences outcomes in these complex cases, so it would be important to know the center-specific breakd