Living donor liver transplantation (LDLT) was initiated in pediatric patients to decrease mortality among patients on the cadaveric donor liver waiting list.1 With excellent patient and graft survival, encouraging results in pediatric LDLT have led to the development of an adult-to-adult LDLT program. As the number of adult LDLTs performed has been increasing, it has been clarified that the size of the graft liver is associated with the clinical outcome. In our previous report,2 when the graft-to-recipient weight ratio (GRWR) was less than 0.8%, the graft survival rate was significantly worse than with larger grafts. The clinical manifestations, referred to as the small-for-size (SFS) syndrome, consist of poor bile production, delayed synthetic function, prolonged cholestasis, and intractable ascites, leading to septic complications and higher mortality. Although a variety of recipient and donor factors are involved in the occurrence of SFS syndrome, persistent portal hypertension or portal venous overperfusion are suggested to be important mechanisms of SFS graft injury. Animal experimental studies3, 4 have shown that partial diversion of portal flow to systemic circulation through a mesocaval shunt can improve the function of an SFS graft liver. A recent clinical report5 also has demonstrated that a mesocaval shunt with downstream ligation of the superior mesenteric vein was effective in preventing SFS syndrome in a recipient transplanted with a SFS liver graft (GRWR of 0.61%). On the basis of these findings, it is plausible that surgical procedures to attenuate portal venous overperfusion might protect a SFS graft from the injuries associated with SFS syndrome. In this article, we describe our new method of using a simple portocaval shunting in LDLT with a SFS graft.
In the development of adult-to-adult living donor liver transplantation (LDLT), the small-for-size graft has been associated with poor clinical outcome. Persistent portal hypertension or portal venous overperfusion are considered to be causative factors, and partial diversion of portal flow to systemic circulation may be effective for avoiding injuries that occur in the small-for-size (SFS) graft. Recently, we constructed an end-to-side portocaval shunting using 1 of the portal branches and anastomosed the other branch with the portal vein of the graft in 2 cases of LDLT recipients transplanted with a SFS graft. With the suppression of portal hypertension, as well as sufficient portal flow to the graft, the recipients recovered successfully with favorable graft function. This new and simple technique may be able to be used as a feasible and effective method to attenuate the SFS syndrome. (Liver Transpl 2004;10:807–810.)
A 16-year-old man collapsed during rugby practice. He was diagnosed with heat stroke, with a core body temperature of 42°C. Because he became comatose and liver function deteriorated markedly, he underwent LDLT for fulminant hepatic failure. He also developed renal failure caused by rhabdomyolysis, which was reflected by a drastic increase in serum creatine phosphokinase concentration to 128,800 IU/L and necessitated hemodialysis before operation. The donor was his mother, who weighed 51 kg; the patient's body weight was 90 kg. A right lobe graft without the middle hepatic vein was transplanted. The graft weight and GRWR were 496 g and 0.55%, respectively. In the recipient, the mean portal venous pressure (PVP) was 24 mm Hg at the time of the insertion of a catheter.6 It increased to 34 mm Hg after clamping of the portal vein (Fig. 1). As a standard procedure to avoid splanchnic congestion during the anhepatic period, a portocaval shunt by end-to-side anastomosis of the right portal branch and the inferior vena cava (IVC) was constructed. After the hepatic venous reconstruction, the small graft was supposed to undergo portal venous overperfusion; therefore, the portocaval shunt was left intact and the right portal vein of the graft was anastomosed to the recipient's left portal branch (Fig. 2). After the reperfusion of the graft with the portal blood, the portal venous pressure was still high (20 mm Hg), when the shunt was occluded and subsequently decreased to 16 mm Hg by opening the shunt. The mean portal vein flow velocity (PFV) of the graft calculated by Doppler ultrasonography decreased from 82 cm/second to 25 cm/second after opening the shunt. Monitoring the PVP in the portal trunk and PFV of the graft, we decided to leave the shunt open (Fig. 1).
Postoperatively, the PVP and PFV transiently increased on day 1, however, they decreased and became stable at between 18 and 22 mm Hg and between 24 and 38 cm/seconds, respectively. The flow through the shunt was also detected by Doppler ultrasonography. Although he required mechanical ventilatory support and continuous hemodiafiltration for a long time after operation, the immediate postoperative graft function was successful. Serum transaminase levels and total bilirubin concentration decreased promptly, and prothrombin time was normalized within 1 week (Fig. 3). On day 9, according to the CT volumetry, the graft volume was estimated at 930 mL, indicating favorable regeneration. Nonetheless, plasma ammonia concentration began to increase at 3 weeks after operation and remained at more than 100 mmol/L for 4 weeks. On day 51, Tc-99m GSA (asialoglycoprotein receptor) scintigraphy revealed a normal value of liver uptake index, which suggested that the shunt flow was minimal. Thereafter, the ammonia level decreased spontaneously. He was weaned from hemodialysis and discharged with satisfactory liver function on day 66.
A 55-year-old man weighing 67 kg underwent LDLT for chronic hepatitis B–related liver cirrhosis and multiple hepatocellular carcinoma. A right lobe graft without the middle hepatic vein was transplanted from his wife, who weighed 49 kg. The graft weight and GRWR were 470 g and 0.70%, respectively. In the recipient operation, a portocaval shunt by end-to-side anastomosis of the right portal branch and the IVC was constructed during the anhepatic period. Judging from the distance between the portal veins of the graft and the recipient, the right portal branch of the recipient was cut off from the IVC and anastomosed to the portal vein of the graft. Then, a new shunt between the left portal branch and the IVC was made. The PFV of the graft was 43 cm/second at the end of the operation; however, it exceeded 100 cm/second despite the shunt being patent during the first 4 days after operation. Afterward, it began to decrease gradually and fell below 50 cm/second within 8 days. The postoperative course was uneventful with a successful graft function (Fig. 4). With no SFS syndrome or hyperammonemia, the patient was discharged on day 25.
Although the pathogenesis of SFS syndrome is multifactorial and has not been clearly identified, a small size of graft is a major contributing factor. Because cadaveric grafts are rare in Japan, LDLT with an SFS graft is sometimes inevitably selected as the only chance for treatment with fully informed consent. Attempts to reduce the risk of SFS syndrome have been made, such as auxiliary transplantation,7 dual liver grafts,8 splenic artery ligation,6 and the prevention of outflow obstruction of the anterior segment in the right lobe graft by reconstruction of additional drainage veins or inclusion of the middle hepatic vein.9 However, from the viewpoint of donor safety, these techniques cannot always be used. Even with these methods, the results have not been satisfactory in some cases. A simple and safe modality to efficiently avoid SFS syndrome is required.
In our previous study, a PVP of more than 20 mm Hg in the early period after LDLT showed a close association with morbidity and poor graft function.6 The current technique can be an option for attenuating the portal hypertension when such an elevated PVP is observed in a patient transplanted with an SFS graft. Compared with the previously reported procedure,5 this is technically simple and should be feasible in most cases. However, there are 2 possible drawbacks with this technique: (1) The portal blood flow competition between the graft and the shunt may cause graft dysfunction because of portal hypoperfusion. (2) The portocaval shunt causes deleterious effects, such as hyperammonemia. However, tactics against both risks can be prepared. If the portal blood is directed predominantly through the shunt, the portal inflow to the graft will be restored by banding the portal branch to the shunt and will be adjusted to the optimal level with the guide of Doppler ultrasonography during the operation. Also, if hyperammonemia is persistent after operation and the symptom is uncontrolled by medical treatment, closure of the shunt can be performed safely as long as graft liver regeneration is confirmed. Because the current technique has been applied and shown to be effective in only 2 cases, the safety and feasibility should be confirmed through accumulation of experience.