Anterior segment congestion of a right liver lobe graft in living-donor liver transplantation and strategy to prevent congestion

Transcription

1 J Hepatobiliary Pancreat Surg (2003) 10:16 25 DOI /s Anterior segment congestion of a right liver lobe graft in living-donor liver transplantation and strategy to prevent congestion SungGyu Lee 1, KwangMin Park 1, Shin Hwang 1, KiHoon Kim 1, ChulSoo Ahn 1, DukBok Moon 1, JungWoo Joo 1, SungHoon Cho 1, KiBong Oh 1, TaeYong Ha 1, HyunSeong Yang 1, KyuTaek Choi 2, KyuSam Hwang 2, EunJoo Lee 2, YoungSang Lee 3, HanJoo Lee 3, YoungHwa Chung 3, MyungHwan Kim 3, SungKoo Lee 3, DongJin Suh 3, and KyuBo Sung 4 1 Department of General Surgery, Division of Hepato-Biliary Surgery and Liver Transplantation, Asan Medical Center, Ulsan University School of Medicine, Poong Nap-dong, Songpa-ku, Seoul , Korea 2 Department of Anesthesiology, Asan Medical Center, Ulsan University, Seoul, Korea 3 Department of Internal Medicine, Asan Medical Center, Ulsan University, Seoul, Korea 4 Department of Radiology, Asan Medical Centre, Ulsan University, Seoul, Korea Abstract Background/Purpose. A left lobe graft from a small donor will not usually fulfill the metabolic demands of a larger recipient in adult-to-adult living-donor liver transplantation (LDLT). One solution to this problem is to use a right lobe graft. However, the necessity of middle hepatic vein (MHV) outflow drainage from the anterior segment (AS) of a right lobe graft has not yet been clearly described in the literature. From July 1997 to February 1998, five right lobe grafts without MHV outflow drainage were implanted in five adult recipients. The graft weights ranged from 650 to 1000 g, and their volumes ranged from 48% to 83% of the ideal liver mass of the recipients. Two grafts showed severe congestion of the AS immediately after reperfusion, followed by prolonged massive ascites and severe liver dysfunction in each patient postoperatively. Eventually, one patient died of sepsis, on posttransplant day 20, demonstrating progressive hepatic dysfunction. Methods. Subsequently, since March 1998, 176 of 208 adult recipients who received a right lobe graft, while demonstrating sizable (greater than 5-mm diameter) MHV tributaries underwent reconstruction of MHV outflow drainage, using the recipient s own autogenous or cryopreserved cadaveric interposition vein grafts. Results. In 170 of the 176 recipients, AS congestion was not demonstrated on enhanced liver computerized tomography (CT) or Doppler ultrasonography (USG) postoperatively, and the patency rate of interposition vein grafts was 96.6% on day 30 posttransplant. Conclusions. A right lobe graft without MHV outflow drainage might result in severe congestion of the AS, which could lead to the patient s death in an extreme situation. Preservation of MHV outflow drainage in a right lobe graft is possible by two harvesting methods: an extended right lobe (ERL) Offprint requests to: S.G. Lee Received: July 18, 2002 / Accepted: July 25, 2002 graft, in which the MHV trunk is included in the graft, and a modified right lobe (MRL) graft, in which venous tributaries of the MHV are reconstructed via interposition vein grafts into the recipient s hepatic venous system. From the viewpoint of donor safety, the ERL graft increases the donor s risk more than the MRL graft, because the remaining left liver lobe of the donor does not possess an MHV. Here, we introduce our experiences of MRL grafts in adult-to-adult LDLTs. Key words Anterior segment congestion of a right lobe graft Modified right lobe (MRL) graft Introduction Living-donor liver transplantation (LDLT) was originally developed as a solution for donor organ shortage in pediatric recipients. 1 However, this procedure has now been widely applied to adult recipients, especially in countries where the availability of brain-dead donors is severely restricted. 2 The major limitation of adultto-adult LDLT is the adequacy of the graft size. 2 To prevent postoperative liver failure, the graft to be implanted should not be too small. Usually, a left lobe graft from a small donor will not fulfill the metabolic demands of a larger recipient. A right lobe graft can become an option to solve the problem of graft-size disparity. However, when a right lobe graft with right hepatic vein (RHV) drainage alone is used in adult-toadult LDLT, severe congestion of the anterior segment (AS) of the right lobe graft can develop in some instances after reperfusion, because the hepatic venous outflow of the AS of the right lobe (subsegments 5 and 8) 3 is drained mostly into the middle hepatic vein (MHV). 4 This AS congestion of the right lobe graft might cause severe graft dysfunction and, further, lead to septic complications and the patient s death in an

2 S.G. Lee et al.: AS congestion in RLLG in living-donor LT 17 Fig. 1A,B. Right liver lobe grafts that have been introduced in living-donor liver transplanation. A Right lobe graft without middle hepatic venous outflow drainage. B Right lobe graft with middle hepatic vein trunk. RHV, right hepatic vein of the liver graft; RHV, right hepatic vein of the recipient; MHV, middle hepatic vein of the liver graft; MHV, middle hepatic vein of the recipient; V5, middle hepatic vein tributary from subsegment 5; V8, middle hepatic vein tributary from subsegment 8 extreme situation. Until now, the necessity of MHV drainage when using a right liver lobe graft has not been clearly described in the literature. 5 8 An extended right lobe liver graft which has an MHV trunk (Fig. 1) avoids this problem and offers better graft function even for a larger recipient. 9 However, the most important ethical issue of LDLT has to involve donor safety. 2 For the donor, LDLT using an extended right lobe (ERL) graft imposes a more extensive operation with a perceived higher risk, even though better survival of the recipient can be anticipated than with a right lobe graft without MHV drainage. Another harvesting method apart from the ERL graft, making possible the preservation of MHV drainage of the right lobe graft, is a modified right lobe (MRL) graft, in which MHV tributaries of the AS are reconstructed via an interposition vein graft into the recipient s hepatic venous system (Fig. 2). Theoretically, from the viewpoint of donor safety, an MRL graft is more advantageous than an ERL graft, because the MRL graft can avoid possible congestion injury of some part of the medial segment of the remaining donor liver by leaving the MHV in the donor and can offer more viable liver tissue in the donor than an ERL graft. Also, an MRL graft provides almost the same functioning liver mass as an ERL graft to the recipient, because all the hepatocytes of the AS are free of venous outflow occlusion and the resultant congestion injury. The purpose of this article is to report our experiences of AS congestion when using a right liver lobe graft without MHV outflow drainage and to introduce the new method of reconstruction of MHV tributaries of the AS by using an MRL graft. Fig. 2. In the modified right lobe (MRL) graft, preservation of middle hepatic venous outflow drainage is carried out by reconstructing the tributaries of the middle hepatic vein via an interposition vein graft into the recipient s hepatic venous system. RHV, right hepatic vein of the liver graft; RHV, right hepatic vein of the recipient; MHV, middle hepatic vein of the recipient; LHV, left hepatic vein of the recipient; V5, middle hepatic vein tributary from subsegment 5; V8, middle hepatic vein tributary from subsegment 8 Experience of anterior segment congestion of right lobe grafts Subjects and methods Among 11 adult-to-adult LDLTs performed from February 1997 to February 1998, five right lobe grafts with only RHV drainage were implanted into five adult recipients. The demographics of the recipients and donors of these five right lobe implantations are listed in Table 1. The indications for LDLT were hepatitis B virus (HBV)-cirrhosis in two, fulminant hepatic failure (FHF) in two, and secondary biliary cirrhosis in one patient. The donors included two wives, one mother, one brother, and one daughter. The body weights of four recipients were larger than those of the corresponding donors, and the lowest donor-to-recipient body weight ratio was 0.78 :1 (donor, 63kg; recipient, 80 kg). Preoperative computerized tomography (CT) volumetry estimated that the graft volume of the left lobe of the donors ranged from 32% to 43% of the standard liver volume (SLV) of the recipients, whereas the values for the right lobe ranged from 51% to 105%. The SLV of the recipient was calculated by our own formula: SLV (ml) 691 body surface (m 2 ) The right lobe graft was chosen because the estimated left lobe volume of the donor was less than 50% of the estimated SLV of the recipient, and the pretransplant recipient s performance was generally poor and categorized as United Network for Organ Sharing (UNOS) status 1 in two patients (FHF as high-urgency

3 18 S.G. Lee et al.: AS congestion in RLLG in living-donor LT Table 1. Demographics of recipients and donors in five adult-adult living-donor liver transplantations using right lobe graft Percentage Anatomical variation No./urgency Recipient disease Donor relation Weight (g) of SLV of the hepatic veins Complications Survival Graft Result 1. Elective Hepatitis B cirrhosis Wife Ascites Alive 2. Urgent Fulminant hepatic Mother Sepsis Alive failure 3. Elective Hepatitis B cirrhosis Brother Massive ascites Alive 4. Urgent Fulminant hepatic Wife Ascites Alive failure 5. Elective Secondary biliary Daughter Hepatic dysfunction, Dead cirrhosis sepsis SLV, standard liver volume; R, right hepatic vein; M, middle right hepatic vein; I, inferior right hepatic vein; 5, hepatic vein of subsegment 5 (V5); 8, hepatic vein of subsegment 8 (V8)

4 S.G. Lee et al.: AS congestion in RLLG in living-donor LT 19 candidates) and 2B in three patients. The estimated remaining left lobe of the donors ranged from 29% to 38% of the whole liver volume of the donors. Donor hepatectomy was performed through an inverted T-incision to the xiphoid process. During the mobilization of a right liver lobe, all sizable ( 5mm) accessory right hepatic veins were preserved. Hilar dissection of the right hepatic artery and right portal vein was done, dividing the portal branch to the caudate process. Intraoperative ultrasound (IOUS) was used to identify the course of the MHV. Temporary clamping of the right hepatic artery and portal vein allowed us to identify the demarcation line. Parenchymal transection was performed along a line 0.5 cm right of the demarcation line, with an ultrasonic dissector and bipolar electrocautery without vascular inflow occlusion. All MHV tributaries of the AS, subsegments 5 and 8 of the right liver lobe (V5 and V8) were divided and ligated during parenchymal transection. In four of the five donors, sizable V5 and/or V8 (range, 9 to 10-mm diameter) were encountered during parenchymal division (Table 1). Therefore, the MHV was kept in the donor s remaining left lobe, and the right lobe graft s venous outflow was dependent upon only an RHV and accessory right hepatic veins if they existed. When a completely prepared right lobe graft maintaining its own vascular flow was still residing in the donor, congestion of the AS did not develop, even though a slight dusky color change was temporarily noticed in one instance. The right lobe grafts were harvested by flushing with 1 l of histidinetryptophan-ketoglutarate (HTK) solution only through the right portal vein cannula. The actual weight of the right lobe grafts ranged from 650 to 1000 g, which corresponded to percentages of the SLV of the recipients ranging from 48% to 83% (Table 1). The recipient operations were performed under active venovenous bypass (Biopump; Medtronic-Bio-Medicus, Minneapolis, MN, USA), except in an adolescent FHF recipient. All accessory right hepatic veins were anastomosed to the side of the recipient s inferior vena cava; both middle and inferior right hepatic veins in two patients and an inferior right hepatic vein in one patient (Table 1). Fig. 3. Postoperative changes in serum aspartate aminotransferase (AST) and serum total bilirubin in case 3 and case 5 Fig. 4. Changes in the amount of drained ascites during the first 3 months in case 3. IVC stent, inferior vena cava stent Results Immediately after portal reperfusion in the recipient, extremely severe congestion and dusky discoloration of the AS of the right lobe graft developed in two (case 3 and case 5) of the five recipients. In case 5, serious bloody oozing from the cut surface of the liver graft after portal venous and hepatic arterial inflow reconstruction was caused by obstruction of the hepatic venous outflow drainage of the swollen AS, and it took another several hours to control the cut-surface bleeding. Immediate postoperative serum aspartate aminotransferase (AST) values were less than 250 IU/dl in cases 1, 2, and 4, but were 780 IU/dl in case 3 and 1706IU/dl in case 5. Case 5 showed a rapidly rising AST value and recorded a maximum of 5497 IU/dl on the first posttransplant day (Fig. 3). Both patients demonstrated evidence of congestive infarcts of the AS on serial CT, checked in the first week posttransplant and the second week posttransplant, respectively. Initial graft dysfunction in case 5 developed into progressive deepening jaundice and sepsis, and led to death on the twentieth day posttransplant (Table 1). Case 3 developed protracted massive ascites ( 10 l per day) from the second week posttransplant and this persisted for a prolonged period (Fig. 4). In case 3, the inferior right hepatic vein was bigger than the right hepatic vein, and was the dominant draining vein of the liver graft. After the congestive infarcts of the AS, the resultant atrophy of the same segment provoked compensatory hypertrophy of the posterior

5 20 S.G. Lee et al.: AS congestion in RLLG in living-donor LT Fig. 5. A In case 3, the inferior right hepatic vein was bigger than the right hepatic vein. B An inferior vena cava stent was installed to control the massive ascites that was caused by compression of the IVC due to the compensatory hypertrophy of the posterior liver segment, which occurred throughout 1 month posttransplant (small black arrows). C After the insertion of a self-made metabllic stent (thick white arrows) through the narrowed segment of the IVC, the interference of the outflow drainage from the dominant inferior right hepatic vein was relieved, and the pressure gradient (from 13 mmhg in the infrahepatic VC to 6mmHg in the suprahepatic VC) disappeared. IVC, inferior vena cava; RHV, right hepatic vein; IRHV, inferior right hepatic vein segment of the graft. This hypertrophied posterior segment progressively compressed the retrohepatic vena cava, and secondarily interfered with the outflow drainage of the inferior right hepatic vein (Fig. 5). Therefore, the mechanism of the prolonged massive ascites in this recipient (maximum, 14 l per day) was the same as that of Budd-Chiari syndrome, that is, it was due to blockage of the dominant inferior right hepatic vein outflow. After the insertion of a self-made metallic expandible vascular stent, the narrowed segment of the inferior vena cava was expanded, and the interference with the outflow drainage of the dominant inferior right hepatic vein was relieved. Finally, the large amount of ascites produced daily was dramatically reduced (Figs. 4 and 5). Case 1 and case 4 also both demonstrated sizable V5 and/or V8 during parenchymal division of the donor liver (Table 1), but in these patients, this resulted in noticeable congestion only in a small part of the AS after graft reperfusion. Postoperatively, these two patients produced a relatively larger amount of ascites than is usual in recipients receiving a cadaveric whole liver graft, but they recovered well. Strategy to prevent congestion of anterior segment After realizing the possibility of the occurrence of AS congestion in the right lobe graft, we used an extended right lobe (ERL) graft, as described by Lo et al. 9 (Fig. 1), in the next three patients. ERL grafts which had an MHV trunk in the graft did not show any evidence of AS congestion, and the early graft functions were satisfactory. However, in one donor, serum total bilirubin rose to 11.3mg/dl in the early postoperative period. In LDLT, donor safety is the most important issue. Therefore, posthepatectomy liver failure in the donor (serum total bilirubin 10 mg/dl) could not be accepted, although she recovered without any sequellae. To ensure both donor safety and satisfactory graft function, we developed a method of MHV reconstruction of the right lobe graft, using an interposition vein graft, which was arbitrarily named, a modified right lobe (MRL) graft (Fig. 2). In the MRL graft, the MHV trunk is left in the donor, and the possible occurrence of congestion in part of subsegment 4 (medial segment) 3 of the remaining donor liver can be avoided. Also, reconstruction of hepatic venous outflow of the AS of the graft, which is performed at the bench surgery, can avoid the possible occurrence of AS congestion injury of the right lobe graft. Technique of preparing and implanting MRL graft During the donor operation, the harvesting method is exactly the same as that for harvesting a right lobe graft, except that before division of the MHV tributaries of the AS (V5 and V8), they are ligated over a short segment with a protective rubber band, or temporarily occluded by applying mini-bulldog vascular clamps to the graft side for future reconstruction. MHV tributaries of the AS which had a caliber of less than 5mm were ligated during liver parenchymal division and most of the MHV tributaries which were reconstructed ranged from 5mm to 15 mm (average, 9.2mm) in diameter. In our first experience of an MRL graft, the long segment

6 S.G. Lee et al.: AS congestion in RLLG in living-donor LT 21 of the autogenous great saphenous vein (GSV) of the recipient was divided into two equal segments. The segments were then split lengthwise and sewn together with 6-0 Prolene suture in order to make the caliber of the vein graft larger. However, this vein graft was lacking in distensibility and elasticity because it had two intervening longitudinal suture lines, and, after reconstruction of V5 and V8, congestion of the AS was obviously noticed and posttransplant graft function rapidly deteriorated. This patient finally needed retransplant, using an ERL graft from another living donor. In the next two successive trials, the recipient s external iliac vein (EIV) was used as an interposition vein graft. Although the EIV graft had functioned satisfactorily, troublesome bloody dissection at the retroperitoneum and resulting hematoma formation at the vein graft harvesting site were major drawbacks of this method, because the recipients had been suffering from chronic end-stage parenchymal liver disease with a poor coagulation profile. Subsequently, a hydrostatically dilated single-lumen GSV, its caliber ranging from 4 to 9 mm (average, 6.2 mm) has been used. From the recipient s left groin, an appropriate length of GSV was harvested and dilated hydrostatically by manual infusion of saline from the caudal end of the vein graft, while the cranial end of the GSV was clamped near its insertion into the femoral vein. A size disparity between the caliber of the GSV and the lumen of the MHV tributaries of the AS (V5 and V8) was usually encountered when the end-toend anastomosis was performed. A longitudinal slit incision was made at the caudal end of the ventral aspect of the GSV lumen, or a branch-patch technique was used when a side branch of the GSV existed, in order to accommodate this size disparity and widen the caudal anastomotic stoma of the GSV with V5 and/or V8. Anastomoses between V5 and/or V8 and the GSV grafts were carried out at the bench surgery, by using a continuous 6-0 Prolene suture. Then, the completed MRL graft was placed in the recipient, and RHV anastomosis was performed first. If the inferior right hepatic vein (IRHV) or the middle right hepatic vein (MRHV) were present in the graft, these accessory right hepatic veins were anastomosed to the side of recipient s retrohepatic vena cava or the corresponding IRHV or MRHV of the recipients, if they existed. After completing the RHV anastomoses, portal vein anastomosis was performed. Then, the liver graft was reperfused by portal blood flow. During the anhepatic phase, Anthrone passive venovenous bypass (Anthrone tube; Toray, Tokyo, Japan) 11 was used for portal inflow diversion into the left femoral vein, instead of the Biopump, since March The GSV interposition vein grafts were temporarily clamped with mini-bulldog vascular clamps and their cephalic ends were anastomosed to the recipient s MHV and/or LHV stumps, in an effort Fig. 6. In this modified right liver lobe graft, three middle hepatic vein tributaries were reconstructed with the interpositioning of an autogenous great saphenous vein, and anastomosed to the recipient s left hepatic vein stump. LHV, left hepatic vein stump of the recipient; V5, middle hepatic vein tributary from subsegment 5; V8, middle hepatic vein tributary from subsegment 8 to create as large an anastomotic stoma as possible to facilitate hepatic vein outflow drainage, by a continuous 6-0 Prolene suture. The reason why an interposition vein graft was anastomosed to the recipient s MHV or LHV instead of the retrohepatic vena cava was to make the outflow orifice of the vein graft being exposed closer to the negative intrathoracic pressure system, and, thereby, to promote hepatic vein outflow drainage more smoothly (Fig. 6). Hepatic artery anastomosis was followed by microvascular surgery. Bile duct anastomosis was performed by Roux-en-Y hepaticojejunostomy or duct-duct anastomosis with internal or external stenting. Statistics The Kaplan-Meier method was used to estimate liver graft survival and the patency rate of the interposition vein graft. Data were statistically analyzed with SPSS for Windows 9.0 (SPSS, Chicago, IL, USA). Results From March 1998 to December 2001, 176 MRL grafts were implanted in 176 adult recipients. The indications for LDLT were hepatitis B virus (HBV) cirrhosis in 104 patients, HBV cirrhosis associated with hepatocellular carcinoma in 32, HCV cirrhosis in 7, Wilson s disease in 2, FHF in 14, retransplant for primary nonfunction of first cadaveric graft in 1, and hepatorenal syndrome (associated with HBV cirrhosis in 13, and alcoholic

7 22 S.G. Lee et al.: AS congestion in RLLG in living-donor LT cirrhosis in 2) in 15 patients. Among the 176 MRL grafts, 1 two-sheet GSV, 2 external iliac veins, and 161 hydrostatically dilated single-lumen GSVs were used as autogenous interposition vein grafts. In the remaining 12 recipients, GSV was not well developed (less than 4 mm in diameter after hydrostatic dilatation) and a cryopreserved cadaveric iliac vein; or the recipient s umbilical and hepatic vein obtained from the resected liver; or the portal vein, ovarian vein, and/or inferior mesenteric vein were substituted as an interposition vein graft. The patency of the interposition vein graft was inspected by Doppler ultrasound every day until posttransplant day 7, and twice a week during the first month posttransplant, and by combined application of magnetic resonance (MR) angiography and threedimensional dynamic CT in some instances. The number of MHV tributaries of the AS that required anastomoses (caliber greater than 5mm) varied from one to four (Fig. 6), but more than three-fourths of the patients had two. Most recipients demonstrated a variable degree of AS congestion when corresponding MHV tributaries were still not reconstructed after portal reperfusion. Also, in many patients, we noticed the outflow of hepatic venous blood from V5 and/or V8 when the mini-bulldog vascular clamps applied to the vein grafts were released. However, after reconstruction of V5 and/or V8 outflow into the recipient s hepatic veins, congestion and discoloration of the AS dramatically disappeared. Postoperatively, the highest serum AST and alanine aminotransferase (ALT) values were less than 250 IU/dl (range, IU/dl) in recipients once the patency of the interposition vein grafts was maintained in the immediate postoperative period. Five patients who received an MRL graft with a hydrostatically dilated single-lumen GSV graft had serum AST of more than 800 IU/dl on the second day posttransplant, and evidence of congestive injury of segment 5 or 8 on follow-up CT (due to occlusion of the interposition vein graft), but they all survived. Of the 176 MRL grafts, 170 demonstrated patency of the interposition vein grafts on serial Doppler ultrasonography, and no evidence of AS congestion injury on dynamic CT examination of the liver graft throughout the follow-up period from 1 week to 12 months after transplantation. Notably, 150 of the 161 recipients that had received MRL grafts with a hydrostatically dilated single-lumen GSV interposition graft are now alive with the original liver graft at a median follow-up of 19.2 months (range, 3 to 36 months). There was no donor mortality. Major postoperative complications in the 176 donors after right hepatic lobectomy were: biliary strictures (n 3), intraabdominal bleeding (n 3), portal vein thrombosis (n 2), and hyperbilirubinemia (n 1; total bilirubin 10 mg/dl). They were successfully managed with medical and surgical treatment. Discussion Although living-donor liver transplantation (LDLT) was originally developed as a viable option for organ shortage in pediatric recipients, this procedure has recently been extended to adult recipients, especially in east Asian countries where the availability of braindead donors is severely restricted. 2 However, extension of LDLT from pediatric recipients to adult recipients has been largely limited because of the inability of a relatively small-size left lobe graft to meet the metabolic demands of a larger adult recipient. It has been postulated that 50% of SLV of the recipient was the minimum graft volume required to provide adequate graft function postoperatively. 12 Also, the use of small-forsize grafts (less than 30% of the recipient s SLV) has been reported to lower graft survival, probably through enhanced parenchymal cell injury, and reduced metabolic and synthetic capacity. 13 Although we previously reported a successful adult-to-adult LDLT using a left lobe graft as small as 30% of the SLV of the recipient in chronic parenchymal liver disease, 14 these small-for-size grafts are more prone to result in postoperative liver failure. Therefore, the major limitation of an adult-toadult LDLT is the adequacy of the graft size. 2 A left lobe from a relatively small donor may not provide a sufficient volume of functioning hepatocytes to the larger recipient. Therefore, LDLT is frequently not possible if the body size of the donor is smaller than that of the recipient, if the implanted liver graft is confined to the left lobe. The solution to this problem is the use of a left lobe plus caudate lobe, auxiliary partial orthotopic liver transplantation, dual-graft implantation, 15 or a right lobe graft. Recently, many institutions have attempted to perform LDLT using a right lobe graft, with varying results. 5 7 It is now perceived that right lobe grafts can help to alleviate the problem of graft size insufficiency in adult-to-adult LDLT. However, when a right lobe graft with RHV drainage alone is used in LDLT, our early experience showed that severe congestion injury of the AS of the right lobe graft developed in two of five patients. According the experimental and clinical experiences of hepatic venous drainage of Nakamura and Tsuzuki, 4 the RHV drains the posterior segment of the right liver lobe and a small part of the AS, particularly the anterior-superior subsegment. In their human autopsy findings, 4 the AS of the right lobe was drained mostly by the MHV. Preparation of a right lobe graft with RHV drainage alone results in division of all draining AS hepatic vein tributaries into the MHV. It was once reported that ligation of the major hepatic veins did not pose any risk in patients undergoing hepatectomy, because an intrahepatic venous collateral circulation developed soon after ligation of the major hepatic veins. 16 We can postulate the mechanisms

8 S.G. Lee et al.: AS congestion in RLLG in living-donor LT 23 Fig. 7A C. Three possible mechanisms by which the hepatic venous outflow obstruction of the anterior segment (AS) of a right liver lobe graft can be tolerated. A First, the thick large right hepatic vein (RHV) primarily drains the area of the AS as well as the posterior segment. B Second, is drainage through the existing interlobar hepatic venous collaterals (HVs) or the new development of collaterals, mainly through sinusoids. C Third, is that the portal vein (PV) of the AS now serves as an outflow tract for the arterial inflow entering the AS. HA, hepatic artery by which the hepatic vein outflow obstruction of the AS can be tolerated (Fig. 7). First, is that the thick right hepatic vein primarily drains the area of the AS, as well as the posterior segment. Second, is that drainage occurs through the existing interlobar hepatic venous collaterals or there is the new development of collaterals, mainly through sinusoids. Third, is that the portal vein of the AS now serves as an outflow tract for the arterial inflow entering the same segment. This retrograde outflow through the portal vein would be the way to decompress the obstructed hepatic venous outflow and to help prevent acute congestion of the AS. However, if instead, the AS is now deprived of nutritional portal inflow, this portal inflow-deprived AS cannot be considered as a functioning part of the right lobe graft in the immediate postoperative period. If intrahepatic venous collaterals between the RHV and the occluded tributaries of the MHV soon develop, the function of the AS will recover. But, if not, atrophy of the AS will develop, due to deprivation of the nutritional portal inflow. In an experimental study, Nakamura et al. 17 clearly indicated that the congestive area due to hepatic vein ligation in the remnant liver cannot be expected to function with the available parenchyma in the early postoperative period. The congestive area resulted in histological necrosis of the hepatic parenchyma 24 h after ligation, although intrahepatic venous collaterals for draining the congestive area were observed through the sinusoids until the seventh day after ligation. They pointed out that the venous drainage through the sinusoids appeared to be insufficient to relieve congestion after hepatic vein ligation in their experiments using rats. The histologically congestive and necrotic area finally became macroscopically atrophic. 17 Although intrahepatic communicating veins are sometimes observed on clinical hepatic venography, it remains controversial whether such intrahepatic collaterals exist in all humans. 4 Therefore, it seems to be logical that the congestive area of the hepatic parenchyma due to hepatic vein ligation cannot be expected to be a functional part of the remnant liver. 17 Considering that a congested AS of a right lobe graft does not have its function, the future function of the right lobe graft has to depend on the remaining posterior segment of the liver graft. The volume ratio between the anterior segment and posterior segment of the right liver lobe in humans has shown considerable variation. In our case 3, the actual weight of the graft was 1000g, equivalent to 83% of the SLV of the recipient. In this patient, the volume of the functioning posterior segment of the graft was greater than 45% of his SLV, and this was enough functioning volume to tolerate the metabolic demands of the recipient in the early posttransplant period. In contrast, the actual weight of the liver graft of case 5 was 770g, equivalent to 76% of her SLV. In this patient, the volume of the functioning posterior segment fell to less than 35% of her SLV, and she was in poor preoperative status (preoperative serum total bilirubin was 47.1 mg/dl without evidence of biliary obstruction) with severe portal hypertension. Thus, it could be understood that both the small functioning graft volume and the poor pretransplant status of the patient, with severe portal hypertension, resulted in graft failure and mortality in case 5. When the right lobe grafts were residing inside the donor, congestion of the AS did not usually develop despite the interruption of hepatic venous drainage of the AS into the MHV. However, after cold storage of the graft, extremely se-

9 24 S.G. Lee et al.: AS congestion in RLLG in living-donor LT vere congestion of the AS of the graft developed in two of five recipients immediately after reperfusion of the liver graft. Mckeown et al. 18 observed that preservation of liver allografts in the cold resulted in cellular swelling, alterations of intercellular organelles, and denudation of the sinusoidal lining. This cold preservation injury to the sinusoidal lining cells, in turn, provoked microcirculatory injury of the graft. The sinusoids in the congested AS of the graft appeared to have a role in the formation of the intrahepatic venous collaterals for draining the congestive area after hepatic vein ligation. 17 However, after cold preservation of the graft, the role of the sinusoids might be compromised and then, the hepatic venous drainage of the AS of the liver graft would deteriorate. This sequence of sinusoidal liningcell injury after cold storage of a right lobe graft might explain the development of AS congestion after reperfusion and the mechanism of the congestive infarction injury. Furthermore, one additional possible mechanism of the AS congestion injury is that portal hypertension exists in recipients with chronic liver disease. In this circumstance, the higher portal inflow and pressure into the AS is an inevitable consequence after reperfusion, and this may further aggravate the hepatic venous congestion of this same segment, which already had colddamaged sinusoids with decreased ability to drain the hepatic venous outflow. At present, we cannot accurately predict the occurrence of AS congestion injury in patients who have received a right lobe graft, because case 1 and case 4 did not show AS congestion injury, even after ligation of 10- mm-caliber V5 and/or V8 (Table 1). The development of AS congestion may depend on whether or not the intrahepatic communicating hepatic vein exists in each individual. In the interest of preventing the possible development of AS congestion and anticipating the recipient s better outcome, preservation of MHV drainage of the AS of the right lobe graft is recommended in any circumstances. Preservation of AS hepatic venous drainage of a right lobe graft is possible by the employment of an extended right lobe (ERL) graft or a modified right lobe (MRL) graft. Although an ERL graft offers better graft function even for a larger recipient, it imposes a more extensive donor operation, with a perceived higher risk. 19 Therefore, from the viewpoint of donor safety, an MRL graft is more advantageous than an ERL graft, because the MHV is left in the donor liver, and this might be able to avoid possible congestion injury of the medial segment of the remaining donor liver. A MRL graft provides almost the same functioning liver mass as an ERL graft, because AS congestion injury of the graft can be avoided by the reconstruction of V5 and V8. Although the diameter of the GSV was smaller than that of external iliac veins, the GSV demonstrated good patency as an interposition vein graft once its diameter was greater than 5 mm after hydrostatic dilatation, and it could be harvested with less invasiveness and a longer segment, if needed, than an EIV. As to the duration of the patency of the saphenous vein graft, we think that 1 to 2 weeks maintenance of the patency is sufficient to avoid congestion injury in the liver graft. The damaged sinusoids of the congestive area of the liver graft will recover to function as intrahepatic venous collaterals in order to drain hepatic venous outflow until the seventh day after operation. 17 Also, continuous regeneration of the liver graft proceeds from immediately after transplantation. Therefore, if obstruction of the interposition vein graft develops at least 1 week after transplantation, AS congestion injury and graft dysfunction probably will not occur, due to the recovery of the sinusoids and the substantially regenerated graft volume. Among our 176 MRL grafts, 170 demonstrated patency of the interposition vein grafts in the immediate postoperative period and no evidence of AS congestion injury in the long-term follow-up. Remarkably, we were able to achieve a 1-year graft survival rate of 92% after adopting the MRL grafts and liberal use of right lobe liver grafts in adult-to-adult LDLT. Our current results show that excellent survival can be expected in adult-toadult LDLT by using our technique when the donor is smaller than the recipient and the estimated left lobe graft volume seems to be inadequate to meet the metabolic demands of the recipient. On the other hand, there are many successful reports of using a right lobe graft without MHV drainge. 5 7 Also, during the same period as the harvesting of the 176 MRL grafts at our institute, another 32 right lobe grafts did not require reconstruction of V5 and V8 because the V5 and V8 exposed during liver parenchymal division were small ( 5-mm-diameter). These 32 right lobe grafts without MHV drainage showed good postoperative function, although 1 of them demonstrated a congestive infarct of subsegment 5 on liver CT. Therefore, controversy may arise regarding whether all right lobe grafts need MHV drainage. If intrahepatic communicating hepatic veins are found, AS congestion will not develop in the graft. As a trial to discriminate the circumstances of future AS congestion in grafts without MHV drainage, we performed preoperative hepatic venography, using a balloon occlusion technique, in five consecutive potential living donors. None of these five donors revealed communication between the RHV and MHV by clinical venography. After portal reperfusion of the MRL graft in the recipients, we have observed not only relatively more severe AS congestion but also stronger outflow of hepatic venous blood from the interposition vein graft of V5 and/or V8, especially in patients with severe portal hypertension, compared with patients with less severe portal hypertension who are recipients.

10 S.G. Lee et al.: AS congestion in RLLG in living-donor LT 25 The question of whether or not the reconstruction of V5 and/or V8 has to be performed in every instance still remains open. So far, there is no good preoperative diagnostic method to predict the development of AS congestion in the right lobe graft of a potential living donor. Simply, we are prudent with the patient who has severe portal hypertension (portal pressure 40 cmh 2 O) who will receive a relatively small right lobe graft (graft volume 50% of SLV), because this circumstance probably entails a higher opportunity to develop AS congestion and graft dysfunction when the reconstruction of large ( 5mm) V5 and/or V8 is not established. In particular, the potential living donor who has a larger anterior segment compared with the posterior segment of the liver, or who has a small RHV has to undergo precise and careful examination to detect the sizable V5 and V8 by preoperative Doppler ultrasonography, multidetector liver CT, and/or MR angiography. Makuuchi 20 recommended the reconstruction of MHV tributaries when the regurgitating blood flow in the liver graft cannot be seen between peripheral tributaries from the MHV to RHV, and also when the portal branch of the AS reveals a retrograde outflow by IOUS after liver division in the donor. During the donor operation, if the discolored area of the AS after the clamping of the right hepatic artery is wide, and the remaining normal-colored area of the right liver graft will be less than 40% of the SLV of the recipient, reconstruction of V5 and/or V8 is required. In east Asia, the scarcity of brain-dead donors is serious, and retransplantation of the liver is not a practical option. Therefore, satisfactory liver graft function should be anticipated at the first operation by providing as many functioning hepatocytes as possible. In LDLT using a right lobe graft, the possible development of AS congestion injury and the resultant graft dysfunction might be prevented by the application of our V5 and/or V8 reconstruction method. References 1. Strong RW, Lynch SV, Ong TH (1990) Successful liver transplantation from a living donor for her son. N Engl J Med 322: Sugawara Y, Makuuchi M (1999) Technical advances in livingrelated liver transplantation. J Hepatobiliary Pancreat Surg 6: Couinaud C (1957) Schema general de la distribution intrahepatique. In: Couinand C (ed) Le foie: etudes anatomiques et chirurgicales. Masson & Cie, Paris, pp Nakamura S, Tsuzuki T (1981) Surgical anatomy of the hepatic veins and the inferior vena cava. Surg Gynecol Obstet 152: Yamaoka Y, Washida M, Honda K, Tanaka K, Mori K, Shimahara Y (1994) Liver transplanation using a right lobe graft from a living related donor. Transplanation 57: Wachs M, Bak T, Karrer F (1998) Adult living donor liver transplanation using a right hepatic lobe. Transplanation 66: Marcos A, Fisher RA, Ham JM (1999) Right lobe living donor liver transplanation. Transplanation 68: Marcos A (2000) Right lobe living donor liver transplanation. Liver Transplanation 6:S59 S63 9. Lo CM, Fan ST, Liu CL (1997) Adult-to-adult living donor liver transplanation using extended right lobe grafts. Ann Surg 226: Hwang S, Lee SG, Lee YJ (1997) Calculation of standard liver volume of Korean adults (in Korean). Korean J HBP Surg 1: Nakao A, Nonami T, Harada V, Kasuga T, Takagi H (1990) Portal vein resection with a new antithrombogenic catheter. Surgery 108: Emond JC, Renz JF, Ferrell LD (1996) Functional analysis of grafts from living donors. Implications for the treatment of older recipients. Ann Surg 224: Kiuchi T, Kasahara M, Uryuhara K (1999) Impact of graft size mismatching on graft prognosis in liver transplanation from living donors. Transplanation 67: Park KM, Lee SG, Lee YJ (1999) Adult-to-adult living donor liver transplanation at Asan Medical center. Transplant Proc 31: Lee SG, Hwang S, Park KM (2001) An adult-to-adult living donor liver transplant using dual left lobe grafts. Surgery 129: Ou QJ, Hermann RE (1987) Hepatic vein ligation and preservation of liver segments in major hepatic resections. Arch Surg 122: Nakamura S, Sakaguchi S, Hachiya T (1993) Significance of hepatic vein reconstruction in hepatectomy. Surgery 114: Mckeown CMB, Edwards V, Phillips MJ, Harvey PRC, Petrunka CN, Strasberg SM (1988) Sinusoidal lining cell damage: the critical injury in cold preservation of liver allografts in the rat. Transplanation 46: Lo CM, Fan ST, Liu CL, Yong BH, Chan JK, Wong J (1999) Increased risk for living liver donors after extended right lobectomy. Transplant Proc 31: Makuuchi M (2000) Overview of standard surgical techniques: left lobe-living donor and split liver transplantation (abstract). The 2nd International Symposium Dedicated to Expanding the Donor Pool; August 26, 2000; Rome, Italy