Left lobe adult-to-adult living donor liver transplantation: Should portal inflow modulation be added?§

Authors


  • Seiji Kawasaki and Yoichi Ishizaki contributed to the study conception and design. Hiroyuki Sugo, Jiro Yoshimoto, and Noriko Fujiwara contributed to the acquisition of data. Seiji Kawasaki, Yoichi Ishizaki, and Hiroshi Imamura contributed to the analysis and interpretation of data. Yoichi Ishizaki and Seiji Kawasaki drafted the manuscript.

  • This research was partially supported by the Japanese Ministry of Education, Culture, Sports, Science, and Technology through a 2006 Grant-in-Aid for Science Research B (18390370). This study was also partially supported by the Japanese Ministry of Health, Labor, and Welfare through a Health and Labor Sciences Research Grant for research on intractable diseases and portal hemodynamic abnormalities.

  • §

    See Editorial on Page 270

Abstract

Recently, the successful application of portal inflow modulation has led to renewed interest in the use of left lobe grafts in adult-to-adult living donor liver transplantation (LDLT). However, data on the hepatic hemodynamics supporting portal inflow modulation are limited, and the optimal portal circulation for a liver graft is still unclear. We analyzed 42 consecutive adult-to-adult left lobe LDLT cases without splenectomy or a portocaval shunt. The mean actual graft volume (GV)/recipient standard liver volume (SLV) ratio was 39.8% ± 5.7% (median = 38.9%, range = 26.1%-54.0%). The actual GV/SLV ratio was less than 40% in 24 of the 42 cases, and the actual graft-to-recipient weight ratio was less than 0.8% in 17 of the 42 recipients. The mean portal vein pressure (PVP) was 23.9 ± 7.6 mm Hg (median = 23.5 mm Hg, range = 9-38 mm Hg) before transplantation and 21.5 ± 3.6 mm Hg (median = 22 mm Hg, range = 14-27 mm Hg) after graft implantation. The mean portal pressure gradient (PVP − central venous pressure) was 14.5 ± 6.8 mm Hg (median = 13.5 mm Hg, range = 3-26 mm Hg) before transplantation and 12.4 ± 4.4 mm Hg (median = 13 mm Hg, range = 1-21 mm Hg) after graft implantation. The mean posttransplant portal vein flow was 301 ± 167 mL/minute/100 g of liver in the 38 recipients for whom it was measured. None of the recipients developed small-for-size syndrome, and all were discharged from the hospital despite portal hyperperfusion. The overall 1-, 3-, and 5-year patient and graft survival rates were 100%, 97%, and 91%, respectively. In conclusion, LDLT with a left liver graft without splenectomy or a portocaval shunt yields good long-term results for adult patients with a minimal donor burden. Liver Transpl 18:305–314, 2012. © 2012 AASLD.

When adult-to-adult living donor liver transplantation (LDLT) was first introduced, left lobe LDLT was the only available option because of the potential risk to the donor posed by right lobe LDLT.1, 2 One of the major problems of adult-to-adult LDLT is that the graft volume (GV) can be insufficient for the recipient's metabolic demands, and left lobe grafts from living donors have been reported to result in small-for-size (SFS) syndrome.3-5 Right lobe LDLT was developed to overcome this graft size problem, and it currently accounts for the majority of adult-to-adult LDLT procedures; the use of left lobe grafts for adults is severely limited by the size limitation.6-9 However, right lobe LDLT poses significant risks to the living donor, including death and substantial morbidity, and 2 highly publicized donor deaths due to right lobe LDLT are thought to have contributed to a decrease in the enthusiasm for this operation.10-13 At present, less than 5% of all liver transplants in the United States involve living donors14; in Japan, however, 99% of liver transplants involve living donors because the availability of brain-dead donors is severely restricted. The crucial prerequisite for LDLT is a minimal risk of morbidity and mortality for the healthy living donor. Previously, we reported that adult-to-adult LDLT should be limited to left lobe grafts; we argued that right lobe donation involves unacceptable risk.2 Because portal hyperperfusion has been reported to be one of the important etiological factors of SFS syndrome, there has been a recent trend in left lobe LDLT to employ portal inflow modulation with techniques such as splenic artery ligation,15-17 splenectomy,16, 18, 19 and portosystemic shunting20-25 to prevent SFS syndrome. However, it is still unclear whether more portal inflow modulation is needed in addition to the existing spontaneous systemic variceal shunts. Without splenectomy or portocaval shunts in any patients, we have been performing adult-to-adult LDLT with left lobe grafts with preoperatively estimated GV/recipient standard liver volume (SLV) ratios ≥ 30%. Here we report our experience with adult-to-adult LDLT with left lobe grafts.

Abbreviations:

GRWR, graft-to-recipient weight ratio; GV, graft volume; HAF, hepatic artery flow; HCC, hepatocellular carcinoma; LDLT, living donor liver transplantation; MELD, Model for End-Stage Liver Disease; PVF, portal vein flow; PVP, portal vein pressure, SFS, small for size; SLV, standard liver volume.

PATIENTS AND METHODS

Patients

Between September 2003 and March 2011, 54 consecutive LDLT procedures were performed at Juntendo University Hospital after approval was obtained from the ethics and indications committee of Juntendo University. The recipients included 42 adults (≥18 years old) and 12 children (<18 years old). All 42 consecutive adult patients underwent LDLT with a left lobe graft without the caudate lobe. These 42 adult recipients and their graft donors were the subjects of this study. The study protocol was approved by the medical ethics committee of Juntendo University School of Medicine, and the study was performed in accordance with the ethical standards established in the 1964 Declaration of Helsinki.

Graft Selection Criteria

We used only left lobe grafts for adult recipients, and we did not perform right lobe LDLT because there is no doubt that right hepatic lobectomy places a serious burden on the donor. The recipient SLV was calculated according to the formula of Urata et al.26 The GV was calculated with computed tomography volumetric analysis, and the actual GV was measured on the back table. Our general selection criteria for grafts in adult-to-adult LDLT included a preoperatively estimated GV/SLV ratio equal to or greater than 30%.

Surgical Technique

The donor hepatectomy (whole left lobectomy including the middle hepatic vein) and the recipient operation were performed as described previously.27 Briefly, the left and middle hepatic veins in the recipient were used for venoplasty. If it was needed, hepatic graft venoplasty was also performed with the left and middle hepatic veins. From a donor whose segment 6 drained through the middle hepatic vein, a whole left graft without the middle hepatic vein was harvested. The left medial vein draining segment 4 joined the middle hepatic vein, and this vein and the left hepatic vein were divided separately. Venoplasty was then performed on the back table to form a common trunk for the drainage vein of the graft in this donor. In all recipients, end-to-end anastomosis of the hepatic veins was performed, the left portal vein in the donor was anastomosed to the left portal vein in the recipient, and the graft was reperfused. The left hepatic artery in the donor was anastomosed to either the right or hepatic artery in the recipient (according to the size match) under direct observation with an operating microscope. Thirty-six LDLT procedures were performed without modification of the graft portal inflow with techniques such as splenectomy and portosystemic shunting, and 6 patients with marked splenomegaly underwent splenic artery ligation because of severe thrombocytopenia. Bile duct reconstruction was performed in a Roux-en-Y fashion. A jejunostomy tube was placed in the afferent limb for intrabowel decompression.

Measurements of the Hepatic Hemodynamics

Intraoperative blood flow measurements were taken with an ultrasonic transit time flow meter (Transonic System, Ithaca, NY) in both the donor and the recipient. The left portal vein flow (PVF) and left hepatic artery flow (HAF) of the donor were measured before dissection of the liver parenchyma was started. In the recipient, after the anastomosis of all the vessels and 15 minutes of equilibration but before biliary reconstruction, the PVF and HAF were measured. The portal vein pressure (PVP) of the native diseased liver was measured before hepatectomy and also after graft implantation by direct puncture with a 25-gauge needle and pressure tubing attached to a normal central venous pressure monitoring transducer. All patients underwent central venous pressure monitoring. The portal pressure gradient was defined as the PVP minus the central venous pressure.

Postoperative Care

The initial immunosuppressive regimen consisted of tacrolimus and prednisone. An intensive anticoagulant treatment, which was performed for more than 2 weeks after LDLT, included the administration of low-molecular-weight heparin at a low dose, antithrombin III, and fresh frozen plasma containing proteins C and S and hematocrit maintenance within the range of 20% to 30%. The postoperative production of ascites, which was removed with indwelling drains, was balanced by an infusion of fresh frozen plasma according to the protein level of the ascitic fluid.

Definition of SFS Syndrome

To evaluate the impact of SFS grafts on outcomes, we used SFS syndrome, which we defined according to both the Kyushu University criteria and the Clavien criteria. According to the Kyushu University criteria, the diagnosis was based on both prolonged functional cholestasis and intractable ascites28 (ie, a total bilirubin level > 10 mg/dL on postoperative day 14 without any other definitive cause of cholestasis and an average daily production of ascites > 1000 mL up to postoperative day 14). According to the Clavien criteria, SFS syndrome was defined as the dysfunction of a small partial liver graft [graft-to-recipient weight ratio (GRWR) < 0.8%] during the first postoperative week after the exclusion of other causes.5 Dysfunction of a partial liver graft was defined as the presence of 2 of the following parameters on 3 consecutive days: a total bilirubin level > 5.8 mg/dL (100 μmol/L), an international normalized ratio > 2, and encephalopathy grade 3 or 4. SFS nonfunction was defined as the failure (graft loss, patient death, or retransplantation) of a small partial liver graft (GWRW < 0.8%) during the first postoperative week after the exclusion of other causes.

Statistical Analysis

Continuous variables were expressed as means and standard deviations or as medians and ranges, and the statistical analysis of continuous variables was performed with a 1-way analysis of variance. Upon the detection of a significant increase by the analysis of variance, post hoc pairwise comparisons were conducted with Tukey's test. Categorical variables were compared with the χ2 test. The statistical analysis of hemodynamic data was performed with a paired sample t test. Calculations were performed with statistical software (JMP, SAS Institute, Cary, NC). Differences with P values < 0.05 were considered to be statistically significant.

RESULTS

Donor and Recipient Characteristics

Detailed demographic data for the recipients and donors are presented in Table 1. The median age of the recipients was 55 years (range = 22-68 years), and 15 were men. The underlying diseases necessitating liver transplantation were hepatitis B or C virus–related cirrhosis (24 cases), alcoholic cirrhosis (3), liver cirrhosis of an unknown etiology (1), primary biliary cirrhosis (6), biliary atresia (3; 1 patient was positive for hepatitis C virus), Alagille syndrome (1), primary sclerosing cholangitis (1), adult-onset type II citrullinemia (1), epithelioid hemangioendothelioma (1), and fulminant hepatic failure (1). Eighteen of the 42 patients had hepatocellular carcinoma (HCC), and 4 had HCC beyond the Milan criteria. Eighteen of the 42 recipients had a history of previous abdominal surgery [appendectomy (5), Kasai procedure (3), splenectomy (2), right hepatic lobectomy (1), distal pancreatectomy with splenectomy (1), gastrectomy (1), cesarean section (1), reduction of intussusception (1), repair of an umbilical hernia (1), hemostasis for an ectopic pregnancy (1), and cholangiographic procedure (1)]. Ten had undergone previous ablation therapy for HCC. Between patients receiving liver grafts that were <0.8% or ≥0.8% of their body weight, there were no significant differences with respect to the recipient age or sex, the etiology, the number of patients with HCC, the Child-Pugh classification, or the Model for End-Stage Liver Disease (MELD) score (Table 1).

Table 1. Recipient and Donor Demographics
Donor DemographicsValueRecipient DemographicsValueActual GRWR < 0.8% (n = 17)Actual GRWR ≥ 0.8% (n = 25)P Value
  • *

    The data are presented as medians and ranges.

Age (years)*34 (20-62)Age (years)*55 (22-68)55 (22-68)56 (27-66)>0.99
Sex: male/female (n/n)33/9Sex: male/female (n/n)15/276/119/160.96
Blood type [n (%)] Etiology [n (%)]    
 Identical29 (69) Viral hepatitis24 (57)11130.85
 Compatible13 (31) Alcoholic3 (7)12 
 Incompatible0 (0) Immune disease7 (17)25 
Relationship with the recipient [n (%)]  Pediatric liver disease4 (10)22 
 Parent2 (5) Other4 (10)13 
 Child25 (59)HCC [n (%)]18 (43)8100.65
 Sibling8 (19)Child-Pugh classification [n (%)]    
 Spouse7 (17) A5 (12)230.96
   B9 (21)45 
   C28 (67)1117 
  MELD score [n (%)]    
   <104 (10)130.39
   10-2027 (64)1314 
   >2011 (26)38 

Liver Grafts

The overall mean actual GV was 432 ± 79 g (median = 420 g, range = 280-690 g), which was equivalent to 39.8% ± 5.7% (median = 38.9%, range = 26.1%-54.0%) of the recipient SLV. The mean actual GRWR was 0.82% ± 0.13% (median = 0.81%, range = 0.48%-1.18%). The actual GV/SLV ratio was less than 40% in 24 of the 42 cases, and the actual GRWR was less than 0.8% in 17 of the 42 cases (Fig. 1). In the 6 patients who underwent splenic artery ligation, the actual GV/SLV ratios were 45.0%, 41.0%, 39.0%, 38.4%, 37.8%, and 36.3%, and the actual GRWRs were 1.03%, 0.84%, 0.81%, 0.80%, 0.79%, and 0.71%, respectively.

Figure 1.

Variability in the size of the actual grafts for the adult recipients (n = 42). The mean GV/SLV ratio was 39.8% ± 5.7% (median = 38.9%, range = 26.1%-54.9%). The mean GRWR was 0.82% ± 0.13% (median = 0.81%, range = 0.48%-1.18%). The GV/SLV ratio was less than 40% in 24 of the 42 cases, and the GRWR was less than 0.8% in 17 of the 42 cases.

Donor Surgical Data and Postoperative Morbidity

The median intraoperative blood loss was 410 mL (range = 70-1080 mL). The median operation time was 548 minutes (range = 380-1106 minutes), and the median vascular clamping time was 91 minutes (range = 0-161 minutes). No donor received an allogeneic blood transfusion during or after surgery. The median remnant liver volume according to volumetric analysis was 869 mL (range = 597-1228 mL), and the median remnant liver volume to total liver volume ratio was 68% (range = 57%-77%). The mean peak postoperative levels of aspartate aminotransferase, alanine aminotransferase, and total bilirubin and the mean prothrombin time/international normalized ratio were 370 ± 234 IU/L, 477 ± 305 IU/L, 1.90 ± 0.93 mg/dL, and 1.34 ± 0.16, respectively. These values were almost normalized at discharge (33 ± 12 IU/L, 58 ± 28 IU/L, 0.59 ± 0.21 mg/dL, and 1.09 ± 0.06, respectively). During the study period, none of the donors died, suffered life-threatening complications, or required reoperation or readmission. Six postoperative complications occurred in 6 donors. Three patients developed gastric volvulus, which required endoscopic correction. Minor biliary leaks developed in 3 donors, but the symptoms were resolved without interventions. The median duration of hospitalization after surgery was 23 days (range = 11-33 days).

Recipient Surgical Data

Table 2 lists intraoperative data for the recipients. The median recipient operation time was 926 minutes (range = 715-1498 minutes), and the median blood loss was 815 mL (range = 75-12,775 mL). The median transfusion volume of packed red blood cells was 0 mL (range = 0-4450 mL), and 30 of the recipients (71%) required no red blood cells perioperatively. The median number of transfused platelet units was 0 (range = 0-35), and 29 patients (69%) required no platelet concentrate. The median volume of transfused fresh frozen plasma was 1795 mL (range = 0-5695 mL). For the comparison of the intraoperative data, the recipients were divided into 2 groups according to their MELD scores (≤20 or >20), but no significant intergroup differences were evident.

Table 2. Recipient Operative Data
 TotalMELD Score > 20 (n = 11)MELD Score ≤ 20 (n = 31)P Value
  • *

    The data are presented as medians and ranges.

Operation time (minutes)*926 (715-1498)907 (807-1380)930 (715-1498)0.65
Anhepatic time (minutes)*62 (38-232)68 (39-161)60 (38-232)0.84
Intraoperative blood loss (mL)*815 (75-12,775)1290 (160-7600)810 (75-12,775)0.57
Red cell blood transfusions [n (%)]12 (29)4 (36)8 (26)0.51
Red cell blood transfusions (mL)*0 (0-4450)0 (0-2205)0 (0-4450)0.86
Platelet transfusions [n (%)]13 (31)3 (27)10 (32)0.76
Platelet transfusions (U)*0 (0-35)0 (0-35)0 (0-20)0.37
Fresh frozen plasma transfusions [n (%)]41 (98)11 (100)30 (97)0.43
Fresh frozen plasma transfusions (mL)*1795 (0-5695)2435 (1350-4050)1750 (0-5695)0.29

Hepatic Hemodynamics

The PVP values were measured for 26 recipients before graft implantation and for 31 recipients after graft implantation. The mean PVP was 23.9 ± 7.6 mm Hg (median = 23.5 mm Hg, range = 9-38 mm Hg) before transplantation and 21.5 ± 3.6 mm Hg (median = 22 mm Hg, range = 14-27 mm Hg) after graft implantation. In 19 of the 31 recipients, the PVP value after graft implantation exceeded 20 mm Hg. The mean portal pressure gradient was 14.5 ± 6.8 mm Hg (median = 13.5 mm Hg, range = 3-26 mm Hg) before transplantation and 12.4 ± 4.4 mm Hg (median = 13 mm Hg, range = 1-21 mm Hg) after graft implantation. In 8 of the 26 patients, the portal pressure gradient after transplantation was higher than 15 mm Hg. In 5 of the 6 patients who underwent splenic artery ligation, the PVP and portal pressure gradient values were measured before ligation and transplantation. The mean PVP and portal pressure gradient values before transplantation did not differ between the patients with splenic artery ligation and the patients without splenic artery ligation [27.2 ± 2.2 versus 23.1 ± 8.2 mm Hg (P = 0.29) and 16.6 ± 3.2 versus 13.8 ± 6.7 mm Hg (P = 0.37), respectively]. The changes in the PVP and portal pressure gradient values for the 25 LDLT recipients whose PVP was measured before and after graft implantation are shown in Fig. 2. The PVP and portal pressure gradient values after graft implantation did not differ significantly from the pretransplant values. The PVP and portal pressure gradient values after graft implantation did not significantly differ from the pretransplant values for patients who received liver grafts that were <0.8% or ≥0.8% of their body weight or for patients who received liver grafts that were <40% or ≥40% of their SLV. Intractable ascites (an average daily production of ascites > 1000 mL up to postoperative day 14) was recognized in 5 of the 8 patients whose portal pressure gradient was > 15 mm Hg and in 8 of the 23 patients whose portal pressure gradient was ≤ 15 mm Hg. The percentage of patients with intractable ascites did not differ between the 2 groups (P = 0.17). In 32 LDLT cases, liver blood flows were measured for both the donor and the recipient. The PVF values after graft implantation were measured for 38 recipients. Although the mean HAF values of the left hepatic artery in the donor and the hepatic artery in the recipient did not differ significantly, the mean recipient PVF value was significantly higher than the mean value of the left portal vein in the donor (Fig. 3A,B). The mean portal contribution to the total hepatic flow in the recipient was significantly higher than the contribution of the left lobe graft in the donor (Fig. 3C). The mean posttransplant PVF value was 301 ± 167 mL/minute/100 g of liver (range = 83-833 mL/minute/100 g of liver). The posttransplant PVF value exceeded 250 mL/minute/100 g of liver in 21 of the 38 recipients. The posttransplant PVP, portal pressure gradient, HAF, and PVF values did not differ between patients receiving liver grafts that were <0.8% or ≥0.8% of their body weight or between patients receiving liver grafts that were <40% or ≥40% of their SLV. Intractable ascites developed in 9 of the 21 patients whose PVF was >250 mL/minute/100 g of liver and in 5 of the 17 patients whose PVF was ≤250 mL/minute/100 g of liver. The percentage of patients with intractable ascites did not differ between the 2 groups (P = 0.39).

Figure 2.

Changes in the PVP and portal pressure gradient values before and after graft implantation (n = 25). The PVP and the portal pressure gradient values after graft implantation did not decrease significantly from the pretransplant values. The data are expressed as means and standard errors of the mean.

Figure 3.

Hepatic hemodynamics of the left lobe before donor resection and after transplantation (n = 32). (A,B) Although the mean HAF values of the left hepatic artery in the donor and the hepatic artery in the recipient did not differ significantly, the mean PVF value of the recipient was significantly higher than the mean value of the left portal vein in the donor. (C) The mean portal contribution to the total hepatic flow in the recipient was significantly higher than the contribution of the left lobe graft in the donor. The data are expressed as means and standard errors of the mean.

Recipient Outcomes

All liver grafts were immediately functional and showed progressive normalization. None of the recipients fulfilled the criteria for SFS syndrome according to the Kyushu University criteria and the Clavien criteria. The smallest graft (310 g or 26% of the SLV) was transplanted into a 62-year-old woman with hepatitis C virus–related end-stage liver cirrhosis, and the recipient's posttransplant course was characterized by immediate graft function and steady normalization of serum total bilirubin and clotting profiles. The average serum aspartate aminotransferase level for the recipients on day 14 after LDLT was 61 ± 47 IU/L, and it returned to less than 100 IU/L in 37 of the 42 patients. In 35 of the 42 patients, the total bilirubin level on day 14 after LDLT returned to less than 3.0 mg/dL despite the small GV/SLV ratio. The average prothrombin time/international normalized ratio on postoperative days 7, 14, 21, and 28 was significantly lower than the pretransplant value in each case. Postoperative complications are summarized in Table 3. No patient developed prolonged cholestasis without a definitive cause for cholestasis. Two patients had prolonged cholestasis (>10 mg/dL) that persisted until postoperative day 14. One of them developed early recurrent hepatitis with a markedly elevated bilirubin level, but this gradually decreased to less than 3.0 mg/dL within several weeks in response to therapy with interferon-α and ribavirin. Although significant ascites production (average daily volume during the first 2 weeks > 1000 mL) was recognized in 17 patients, all 42 patients progressed well, and none had persistent ascites.

Table 3. Recipient Postoperative Course
 TotalActual GRWRP ValueActual GV/SLV RatioP ValueMELD ScoreP Value
<0.8% (n = 17)≥0.8% (n = 25)<40% (n = 24)≥40% (n = 18)>20 (n = 11)≤20 (n = 31)
  1. NOTE: The data are presented as numbers and percentages or as numbers.

SFS syndrome          
 Prolonged cholestasis2 (5)110.78110.84110.46
 Intractable ascites17 (40)6110.57890.285120.70
 Coagulopathy0 (0)000000
Acute rejection8 (19)170.056530.73350.43
Recurrent hepatitis C10 (24)370.43550.60190.15
Acute renal failure3 (7)030.071120.39120.13
Vascular complication          
 Hepatic arterial thrombosis0 (0)000000
 Portal vein thrombosis0 (0)000000
 Outflow block1 (2)010.31010.19010.43
Biliary complication          
 Bile leakage2 (5)200.052110.84110.46
 Biliary stricture2 (5)110.78020.061110.46
Reoperation5 (12)140.30230.41140.73
In-hospital death0 (0)000000

Episodes of rejection were encountered in 8 patients, but each was treated successfully with steroid bolus therapy and/or OKT3. Ten patients developed early recurrent hepatitis C, but they were treated successfully with the administration of interferon and ribavirin. There were no episodes of arterial or portal vein thrombosis during the postoperative course in any of the 42 patients. Only 1 patient developed a late-onset venous outflow obstruction, which was treated by refixation of the round ligament and placement of a Foley balloon catheter into the right subphrenic space.29 Two patients who developed biliary strictures required transhepatic biliary drainage via laparotomy (on postoperative days 33 and day 80, respectively), and they were cured successfully by balloon dilation via the drainage route.

There was no significant difference in the rate of posttransplant complications between the high-MELD group (MELD score > 20) and the low-MELD group (MELD score ≤ 20; Table 3). There was no difference in the incidence of posttransplant complications between patients receiving liver grafts that were <0.8% or ≥0.8% of their body weight or between patients receiving liver grafts that were <40% or ≥40% of their SLV (Table 3). The overall 1-, 3-, and 5-year patient and graft survival rates were 100%, 97%, and 91%, respectively, within a median follow-up period of 38 months (Fig. 4). None of the recipients died within 2 years of LDLT. One patient died 2.2 years after LDLT because of a mitral valve insufficiency, and another died 4.2 years after LDLT because of liver failure related to recurrent hepatitis C virus disease.

Figure 4.

Overall patient and graft survival for adult-to-adult LDLT. The overall 1-, 3-, and 5-year patient and graft survival rates were 100%, 97%, and 91%, respectively, within a median follow-up of 38 months. None of the recipients died within 2 years of LDLT.

DISCUSSION

Since the first successful LDLT with a left lateral segment graft, which was performed in a small child by Strong et al. in 1990,30 the extension of the use of left lobe liver grafts in LDLT to large adolescents and adults has resulted in satisfactory graft and patient survival outcomes.2, 31 However, the term SFS syndrome was coined to describe significant functional impairment in some recipients of small grafts; this impairment is characterized by prolonged cholestasis, ascites, coagulopathy, encephalopathy, and histological changes consistent with ischemic injury and often results in graft failure or retransplantation.3-5, 32, 33 In general, the survival of liver grafts with a GV/SLV ratio < 40% or a GRWR < 0.8% has been found to be poor. Kiuchi et al.3 reported that in elective cases, the actual graft survival rates at 1 year were 42%, 74%, and 92% when the GRWRs were <0.8%, 0.8% to 1.0%, and 1.0% to 3.0%. Ben-Haim et al.4 reported that when grafts with a GRWR < 0.85% were used in Child B or C patients, the graft survival rate was only 33%. Sugimoto et al.32 reported that 3 of 8 recipients of small grafts (GV/SLV ratio < 40%) died because of graft failure caused by hyperperfusion. Sugawara et al.33 suggested that the hospital mortality rate of recipients with a GV/SLV ratio ≤ 40% was significantly higher than the rate of recipients with a GV/SLV ratio > 40% (21% versus 4%, P = 0.02), and a GV/SLV ratio ≤ 40% provided a lower chance of survival after adult-to-adult LDLT. However, in the present series, although the sample size was small, none of the 42 patients who received a left lobe graft fulfilled the criteria for SFS syndrome, and the 1-year graft and patient survival rates were both 100%.

LDLT poses significant ethical tension. It has the potential to substantially increase the number of livers available for transplantation and, therefore, decrease the overall mortality rate for candidates awaiting transplantation. However, this benefit must be weighed against the risks of morbidity and mortality borne by the healthy volunteer donor. Donor deaths after right lobectomy for adult-to adult LDLT have been reported in Japan, the United States, and Europe. Some of these donor deaths were secondary to liver insufficiency after right lobectomy.11-13 The Japanese Liver Transplantation Society previously reported that the frequency of complications was significantly higher for right liver lobe donors versus lateral segment and left lobe donors.34 Taketomi et al.35 reported that the mean peak postoperative total bilirubin level and the duration of the hospital stay after LDLT were significantly less for donors of left lobe grafts versus donors of right lobe grafts. It is indisputable that right lobectomy carries a risk of residual liver insufficiency (albeit a very low one), whereas this is very unlikely after left lobectomy.

In adult-to-adult LDLT, donor selection limits the overall utilization of this technique. In the present series, 19 of the 61 potential donors were excluded from the donor evaluation. Nine of these 19 excluded donors were found to be unacceptable for left lobe donation because the left lobe graft was considered to be too small for the recipient. Because the donor's safety takes precedence over the recipient's outcome, the remaining liver volume of the donor should not be less than 30% of the total liver volume. If a donor candidate is rejected because the left lobe volume is less than 30% of the recipient's SLV, the donor has a relatively large right lobe that is suitable as a graft for an adult recipient. However, after the donation of the right lobe, the donor's remaining left lobe would probably be too small to ensure his or her own maximum safety, so donation of either the left or right liver lobe would be unacceptable.36, 37

One of the major causative factors of SFS syndrome is assumed to be portal hyperperfusion.5, 38, 39 We have shown previously that after the reperfusion of a left lobe graft, the hemodynamic changes are much more pronounced than those occurring in cadaveric donor liver transplantation.27 The impact of portal vein inflow on the development of SFS syndrome has led to the development of several techniques to decrease portal venous hyperperfusion, including splenic artery ligation,15-17 splenectomy,18, 19 and hemiportocaval shunting.20-25 Troisi et al.23 performed 13 adult-to-adult LDLT procedures with a GRWR < 0.8% and compared 5 patients without any portal inflow attenuation to 8 patients with graft inflow modulation via a hemiportocaval shunt. They found that SFS syndrome occurred in 3 patients without graft inflow modulation but in none of the patients in the graft inflow modulation group. Botha et al.25 reported a series of 21 patients undergoing left lobe adult-to-adult LDLT in 2 institutions; 16 received a hemiportocaval shunt. They suggested that the transplantation of small grafts with modulation of the portal inflow through the creation of a hemiportocaval shunt might prevent the development of SFS syndrome. Yamada et al.24 suggested that a PVP > 20 mm Hg after graft implantation should be lowered with a hemiportocaval shunt to prevent graft failure. However, a hemiportocaval shunt could in itself constitute a risk of hepatofugal flow, which could in turn lead to graft dysfunction, liver atrophy, or portal thrombosis. Moreover, a hemiportocaval shunt may excessively divert the portal flow into the systemic circulation and create a risk of encephalopathy. Botha et al. reported that 2 patients in their series who developed encephalopathy required closure of the shunt.

PVP values ≤ 15 to 20 mm Hg17, 19, 24 and PVF values ≤ 250 to 260 mL/minute/100 g of liver18, 38, 40 have been reported to be suitable target levels for the prevention of SFS syndrome. Ogura et al.19 reported that the 1- and 2-year survival rates of patients with a PVP ≥ 15 mm Hg were 73.0% and 66.3%, respectively. Troisi et al.38 reported that in 5 cases involving SFS grafts (GRWR < 0.8%) with PVF values > 250 mL/minute/100 g of liver, 3 patients without graft inflow modulation developed SFS syndrome and required retransplantation.38 In the present series, the PVP values after graft implantation were >20 mm Hg in 19 of the 31 recipients for whom this value was measured, and the posttransplant PVF value was >250 mL/minute/100 g of liver in 21 of the 38 recipients for whom this value was measured. Even though these cases had excessive portal blood perfusion and neither portocaval shunting nor splenectomy was performed in any of the cases in our series, none of the patients developed SFS syndrome. There is currently no definite consensus on the optimum PVP or PVF value in terms of subsequent patient or graft survival after LDLT.

In our series, there was no difference in the PVP or PVP gradient before and after transplantation. Theoretically, the size of a partial liver graft is related to the magnitude of the PVP change: smaller grafts are assumed to cause an elevation of the PVP after graft implantation. On the other hand, a grafted liver that has a normal texture is considered to have better compliance than the native cirrhotic liver. These situations associated with left lobe grafting are thought to explain why neither the PVP nor the portal pressure gradient was changed after graft implantation. Although in approximately 60% of the patients the PVP value after graft implantation was >20 mm Hg, the portal pressure gradient after transplantation was higher than 15 mm Hg in 8 of the 26 patients (31%). This discrepancy could be attributable to a difference in the central venous pressure. Botha et al.25 reported that in 11 patients who underwent LDLT with a left lobe graft, the portal pressure gradient was reduced from a median of 18 to 5 mm Hg after the creation of a hemiportocaval shunt. An excessively high portal pressure gradient may lead to SFS syndrome, and in these patients, portal inflow modulation may be necessary. Further experience and data are needed to clarify this issue.

It is generally accepted that the recipient's condition is an important variable because a larger GV is necessary when the recipient's pretransplant liver function is severely impaired.4, 41, 42 Yi et al.43 reported that patients with a high MELD score who received an SFS graft had lower 1-year survival rates than patients who received a normal size graft (71.7% versus 86.3%). Ikegami et al.44 also admitted that among patients with a high MELD score (>25), the 1-year survival rate for those receiving liver grafts with an SLV < 35% was only 57.1%.44 In the present series, all recipients with a MELD score > 20 were discharged from the hospital. Although extreme caution is required for very sick recipients of SFS grafts, an accumulation of experience means that adult-to-adult LDLT with SFS grafts can result in good graft function and survival rates, even in recipients with high MELD scores.

Having previously emphasized the safety of right lobe grafts, several groups have changed their concept of graft selection for adult-to-adult LDLT.15, 17, 22 Since the early stage of adult-to-adult LDLT development, we have reported that a left lobe graft can maintain good viability and regeneration without portal inflow modulation.2, 27 LDLT with a left lobe graft has yielded good short- and long-term results for adult-to-adult patients with a minimal burden on the donors. Left lobe grafting without PVF modulation may be more frequently applicable than previously supposed, especially in patients whose posttransplant portal pressure gradient is not too high and who have been able to receive a liver graft that is ≥30% of their SLV.

Ancillary