For patients with diffuse thrombosis of the splanchnic venous system, portal inflow via cavoportal hemitransposition has been proposed in deceased donor whole liver transplantation.[1-5] In living donor liver transplantation, cavoportal hemitransposition is rarely indicated because a partial liver graft cannot tolerate the high systemic blood flow through the inferior vena cava. In some patients, a large pericholedochal varix draining from pancreatic collaterals can be a useful vessel for providing splanchnic blood flow with hepatotrophic factors to the graft and for relieving portal hypertension. Two adult recipients with diffuse thrombosis were successfully treated with a large pericholedochal varix for portal inflow.
In total portosplenomesenteric thrombosis patients, cavoportal hemitransposition (CPHT) is indicated but rarely applicable for adult-to-adult (A-to-A) living donor liver transplantation (LDLT) because partial liver graft requires splanchno-portal inflow for liver graft regeneration. If intra- & peri-pancreatic collaterals draining into pericholedochal varix were present, pericholedochal varix may provide splanchnic blood flow to the transplanted liver and also relieve recipient's portal hypertension. To date, however, there is no successful report using pericholedochal varix in liver transplantation (LT). We successfully performed A-to-A LDLTs using pericholedochal varix for those 2 patients. The surgical strategies are followings: (a) dissection of hepatic hilum to isolate left hepatic artery using for arterial reconstruction of implanted right lobe graft, (b) en-mass clamping of the undissected remaining hilum if we can leave adequate length of stump from the clamping site, and then hilum is divided, (c) delay the donor hepatectomy until the feasibility of the recipient operation is confirmed. Portal flow was established between the sizable pericholedochal varix (caliber > 1cm) and graft portal vein, but the individually designed approaches were used for each patients. Currently, they have been enjoying normal life on posttransplant 92 and 44 months respectively. In conclusion, enlarged pericholedochal varix in patients with totally obliterated splanchnic veins might be an useful inflow to restore portal flow and secure good outcome in A-to-A LDLT. Liver Transpl 20:612–615, 2014. © 2014 AASLD.
Patient 1 was a 50-year-old man with hepatitis B virus–related cirrhosis, ascites controlled by diuretics, and a Model for End-Stage Liver Disease score of 21 points. A preoperative computed tomography scan revealed diffuse splanchnic thrombosis and a pericholedochal varix more than 1 cm in diameter (Fig. 1A). The following surgical strategies were considered: (1) a trial dissection of the hepatic hilum to isolate the left hepatic artery and use it for the arterial inflow of the graft, (2) trial en masse clamping of the undissected remaining hilum if there was an adequate length of the stump distal to the clamp, and (3) a delay of the donor right hepatectomy until the feasibility of the recipient operation was confirmed.
During the exploration, the dissection of the left hepatic artery was relatively easy without much bleeding because the left hepatic artery was located on the medial side of the hepatoduodenal ligament at some distance from the bile duct with the pericholedochal venous plexus. Additional dissection of the hepatic hilum was not performed because of copious bleeding from the pericholedochal venous plexus; the donor hepatectomy was then started. After completion of the recipient's peri- and retrohepatic dissection, the recipient's great saphenous vein was procured for reinforcement of the large, thin-walled choledochal vein and plasty of the recipient's right hepatic vein. After en masse clamping of the hepatoduodenal ligament (except for the isolated left hepatic artery), the recipient and donor hepatectomies were performed simultaneously to reduce undue anhepatic time. We isolated the choledochal varix at the stump of the undissected hepatic hilum, and the remaining structures were closed with 6-0 or 5-0 polypropylene sutures. The choledochal varix was reinforced with matrix sutures via the wrapping of the bisected great saphenous vein in preparation for reconstruction (Fig. 1B-D). The implantation of the modified graft (650 g and 1.1% graft-to-recipient weight ratio) was performed as previously described. The left hepatic artery was used for arterial reconstruction, and Roux-en-Y hepaticojejunostomy was performed.
Patient 2 was a 58-year-old man with hepatitis B virus–related cirrhosis, hepatocellular carcinoma, and a Model for End-Stage Liver Disease score of 18 points. He had a small amount of ascites and esophageal varices without red wale signs. A preoperative computed tomography scan revealed an obliterated portosplenomesenteric vein and a large pericholedochal varix (Fig. 2A). The surgical procedures were similar to those previously described for patient 1. Before the division of the hepatic hilum for the total hepatectomy, however, we isolated the enlarged pericholedochal vein, which was 17 mm in diameter and 4 cm in length and had a thick wall. The weight of the modified right lobe graft was 900 g, and the graft-to-recipient weight ratio was 1.22%. The right portal vein of the graft was fenced with the funnel-shaped bisected great saphenous vein for size matching with the pericholedochal varix on the back table (Fig. 2B,C). For arterial and biliary reconstruction, the left hepatic artery and the jejunal Roux limb, respectively, were used.
Both patients did well after transplantation (92 and 44 months, respectively). Patient 1, however, developed a large amount of ascites and severe stenosis of the portal vein anastomosis site at 5 months, which was successfully treated with percutaneous stent placement in both the right anterior portal vein and the posterior portal vein crossing the anastomotic site.
To the best of our knowledge, this is the first successful report of deceased donor or living donor liver transplantation using a pericholedochal varix. In our experience, first, the pericholedochal varix should be more than 1 cm in diameter. Multiple small-diameter pericholedochal varices cannot provide adequate portal inflow to the implanted graft and may result in graft failure. Both patients had small amounts of ascites without large esophageal varices, and this was indirect evidence of effective decompression of portal hypertension through the enlarged pericholedochal varix. Second, the wall of the enlarged pericholedochal varix is usually very thin and densely attached to the bile duct wall because of periductal fibrosis. Even though it has a thick wall because of long-standing portal hypertension, troublesome bleeding is inevitable during dissection. Proper use of en masse clamping of the hepatic hilum after isolation of the left hepatic artery, however, enabled us to perform the procedures without massive bleeding. Third, it is difficult to dissect the artery because of the enlarged pericholedochal varices, particularly around the right hepatic artery, and the use of the left hepatic artery should be considered.
In conclusion, in a patient with obliteration of the entire splanchnic venous system but without severe esophageal varices and ascites, an appropriately sized pericholedochal varix can provide portal inflow without cavoportal hemitransposition.