By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Few studies have examined the long-term outcomes and prognostic factors associated with pediatric living living-donor liver transplantation (LDLT) using reduced and hyper-reduced left lateral segment grafts. We conducted a retrospective, single-center assessment of the outcomes of this procedure, as well as clinical factors that influenced graft and patient survival. Between September 2000 and December 2009, 49 patients (median age: 7 months, weight: 5.45 kg) underwent LDLT using reduced (partial left lateral segment; n = 5, monosegment; n = 26), or hyper-reduced (reduced monosegment grafts; n = 18) left lateral segment grafts. In all cases, the estimated graft-to-recipient body weight ratio of the left lateral segment was more than 4%, as assessed by preoperative computed tomography volumetry, and therefore further reduction was required. A hepatic artery thrombosis occurred in two patients (4.1%). Portal venous complications occurred in eight patients (16.3%). The overall patient survival rate at 1, 3 and 10 years after LDLT were 83.7%, 81.4% and 78.9%, respectively. Multivariate analysis revealed that recipient age of less than 2 months and warm ischemic time of more than 40 min affected patient survival. Pediatric LDLT using reduced and hyper-reduced left lateral segment grafts appears to be a feasible option with acceptable graft survival and vascular complication rates.
Liver transplantation (LT) is an established curative therapy for children with end-stage chronic liver disease or acute liver failure. Outcomes following LT have significantly improved, due to advances in surgical procedures, preservation technology, immunosuppressive management and perioperative care.
The shortage of organ donors for low-weight LT recipients, especially small children, has led to the development of new surgical techniques to increase the donor pool. To increase the supply of appropriately size organs for pediatric recipients, the techniques of reduced, split and living donor liver transplantation (LDLT) were developed. These technical innovations have expanded the potential donor pool and have led to a significant decrease in waiting list mortality for children (1). Graft-to-recipient weight ratio (GRWR) sometimes exceeds 4.0% for neonates and small infants receiving left lateral segment graft; in such cases, the graft does not fit into the abdomen, due to the problem of a large-for-size graft, and further reduction is required (2,3). Reduced left lateral segment grafts (partial left lateral segment and monosegment grafts) and hyper-reduced left lateral segment grafts (reduced monosegment grafts) can be employed as a solution to this problem.
Available articles in the literature are based on small samples or even case reports, and are deficient in long-term follow-up, making it difficult to draw conclusions about the outcomes of reduced and hyper-reduced left lateral segment grafts. We report our results and long-term follow-up for LDLT using partial left lateral segment, monosegment and reduced monosegment grafts from living donors for neonates and extremely small infants.
Patients and Methods
Of the 295 children under 18 years undergoing LDLT in our center between September 2000 and December 2009, 49 patients, 22 male and 27 female, underwent LDLT using partial left lateral segment (n = 5), monosegment (n = 26), and reduced monosegment grafts (n = 18). Table 1 shows the characteristics of pediatric LDLT with reduced and hyper-reduced left lateral segment grafts. In all cases, the estimated GRWR of the left lateral segment exceeded 4% as estimated by preoperative computed tomography (CT) volumetry, and further reduction was therefore required. Reduction to a partial left lateral segment graft, monosegment graft or reduced monosegment graft varied based on the exact graft volume and shape required. These children had end-stage liver disease and could not delay treatment to grow up and receive conventional LDLT using a left lateral segment graft. All patients received grafts from a parent or grandparent.
Table 1. Characteristics for pediatric LDLT with reduced and hyper-reduced left lateral segment grafts
The median patient age was 7 months (range, 1–16 months), the median body weight 5.45 kg (range, 2.77–8.05 kg) and the median GRWR 3.5% (range, 2.26–5.29%). Ten patients received ABO-incompatible grafts, nine received ABO-compatible grafts and 30 received identical ABO-identical grafts. The mean cold ischemic time (CIT) was 67.8 min (median 50; range, 8–198 min), the mean warm ischemic time (WIT) 46.7 min (median 43; range, 25–92 min), the mean operative time 10.5 h (median 9 h 46 min; range, 7.2–15.7 h), the mean blood loss 953 g (median 640 g; range, 40–6100 g) and the mean preoperative pediatric end-stage liver disease (PELD) score was 27 (median 27; range, 8–50). Pre-LDLT stay was categorized as either in the home, hospital or intensive care unit (ICU). Eight patients (16.3%) were in the ICU at the time of LDLT.
The common indications for pediatric LDLT using reduced and hyper-reduced left lateral segment grafts, were biliary atresia (BA; 63.3%) and fulminant hepatic failure (FHF; 16.3%). Potential donors were evaluated using liver function tests, blood group testing, human leukocyte antigen typing, identification of anatomical variations and graft size. The operative procedure was as previously described (2,4). The hepatic venous anatomy of the donor was examined using both preoperative three-dimensional CT and intraoperative ultrasonography. After isolation of the left donor hepatic artery, hepatic duct and portal branch, the hepatic parenchyma of the medial segment was transected 5 mm to the right of the falciform ligament without blood inflow occlusion or graft manipulation. For the initial ten cases, for the segment III monosegment graft (reduced graft), sterile methylene blue solution (one to ten dilution) was injected into the portal vein (PV) of segment II to allow clear demarcation between two segments as previously reported (5,6). Accordingly, segment II was resected from the left lateral segment with the use of the ultrasonic dissector, in accordance with the technique described by Srinivasan et al. (5). Thereafter, we realized from these experiences that the determination of transection plane between segments II and III was approximately in-between hepatic veins of segment II and III, made by intraoperative ultrasonography (Figure 1: Plane A; Ref. 4). The decision of reduction and further reduction of graft was based largely on experienced surgeon's clinical assessment of the size, shape and thickness of the graft relative to the size of child abdomen. For the reduction of a monosegment to a reduced monosegment (hyper-reduced graft), the caudal portion of the monosegment was resected in situ and discarded (Figure 1: Plane B). In this reduction, the distal portion of the portal branch near segment III was severed (2). In the partial left lateral segment grafts (reduced graft), all grafts were initially left lateral segment grafts, and the parenchymal plane of resection for reduction was a sagittal plane that resected the left lateral edge of the graft (Figure 1: Plane C), as described elsewhere (7).
The hepatic artery reconstruction was performed using a microvascular technique in all cases. Biliary reconstruction was achieved using a Roux-en-Y hepatico-jejunostomy in all cases, except one in which duct-to-duct anastomosis was used instead. In the era prior to April 2006, the PV was reconstructed in the same way as in a pediatric LDLT using a left lateral segment forming what is known as a “C-curve shaped” PV (Figure 2A). Next, we elected to cut the recipient's PV diagonally close to the pancreas and shorten it as much as possible after April 2006 (Figure 2B). The abdominal closure procedures at the time of transplantation were full thickness primary closure in 37 cases, partial primary closure in six cases and skin closure in six cases.
Postoperative management and immunosuppression
During the 2 weeks after the LT, management included daily blood tests and Doppler ultrasonography. To avoid thrombotic events, intravenous heparin was administered for the first 7 days maintaining the activated partial thromboplastin time approximately 40 seconds and was subsequently tapered off gradually, and oral antiplatelet aggregation drugs were administered for the first 6 months.
The immunosuppressive regimen consisted of tacrolimus (FK506) and low-dose steroids (8). Target tacrolimus trough levels were initially more than 10 ng/mL, decreasing gradually to 6–8 ng/mL a few months after LDLT. Methylprednisolone (10 mg/kg body weight) was administered intraoperatively before reperfusion. During the postoperative period, 1 mg/kg of the same drug was given for the first three postoperative days followed by 0.5 mg/kg for the next 3 days and 0.3 mg/kg on the seventh postoperative day. Individuals were transitioned to oral prednisolone at a dose of 0.3 mg/kg on the eighth postoperative day. Prednisolone was reduced to 0.1 mg/kg/day 4 weeks after the transplantation if the postoperative course was free of liver dysfunction, and steroid therapy was routinely weaned by 3–6 months after transplantation provided that graft function was maintained.
Acute cellular rejection (ACR) was treated with high dose methylprednisolone as pulse therapy. If rejection was steroid resistant, Muromonab-CD3 was used. If liver function did not clearly normalize, a third drug for maintenance immunosuppression was added, specifically, either azathioprine for earlier cases or mycophenolate mofetil (MMF) for later cases. All episodes of rejection were diagnosed by liver biopsy.
Overall patient survival was described using Kaplan–Meier methods; the log-rank test was used for the univariate analysis of risk factors. The cut-off values for the univariate analyses were determined using minimal p value approach (9). The primary outcome was defined as graft failure or patient death after LDLT and Cox's proportional hazard model was used to examine prognostic factors. Note that p values of less than 0.05 were considered to be significant. JMP version 9 (SAS Institute Inc., Cary, NC, USA) was used for all statistical analyses.
We observed no incidence of donor mortality or morbidity in our study group. The mean blood loss during the donor operation, which is mainly from the surgical margin, was 180 g (median 200 g; range, 70–270 g) in partial left lateral graft, 144 g (median 120 g; range, 35–310 g) in monosegment graft and 218 g (median 220 g; range, 20–540 g) in reduced monosegment graft; the difference was not significant (p = 0.152).
Table 2 shows recipients and graft profiles of partial left lateral (n = 5), monosegment (n = 25) and reduced monosegment (n = 19) grafts. Graft weight was significantly smallest in reduced monosegment (p < 0.001). However there was no significant difference in GRWR, discarded volume or reduction rate.
Table 2. Recipient and graft profiles of reduced and hyper-reduced graft
Recipient body weight (g)
Graft weight (g)
Discarded volume (g)
Reduction rate (%)
GRWR = graft-to-recipient weight ratio.
Kruskal–Wallis test, Dunn's post-test: *p < 0.05 versus reduced monosegment, ***p < 0.001 versus reduced monosegment.
Partial left lateral segment (n = 5)
5859 ± 1100
209 ± 33*
3.65 ± 0.78
69 ± 48
26.3 ± 16.7
Monosegment (n = 25)
6112 ± 953
214 ± 48***
3.74 ± 0.80
91 ± 40
31.4 ± 13.9
Reduced monosegment (n = 19)
4700 ± 1327
152 ± 32
3.37 ± 0.82
112 ± 44
38.7 ± 12.3
Survival outcomes and causes of mortality
At a median follow-up of 64 months (range, 0.7–125.5 months), the actuarial patient survival in reduced or hyper-reduced graft at 1, 3, 5 and 10 years after LDLT (Figure 3) was 83.7%, 81.4%, 78.9% and 78.9%, respectively, and the graft survival at 1, 3, 5 and 10 years after LDLT was 83.7%, 81.4%, 78.9% and 75.2%, respectively. The survival rate of patients with reduced or hyper-reduced graft was lower than patients (n = 246) with left lateral segment graft in the same period (p = 0.033) (Figure 3). Patient survival improved after April 2006 but not significant (p = 0.243) (Figure 4). Ten patients died during the study period (Table 3), two unrelated to complications after LDLT. Sepsis (n = 4) and refractory severe rejection (n = 3) were the major causes of related mortality. One patient died secondary to a hepatic artery thrombosis (HAT).
Table 3. Cases of mortality after pediatric LDLT using reduced and hyper-reduced left lateral segment grafts
Major cause of mortality
Type of graft
FHF = fulminant hepatic failure; IHHE = infantile hepatic hemangioendothelioma; LC = liver cirrhosis; CAPV = congenital absence of portal vein; BA = biliary atresia; HAT = hepatic artery thrombosis.
4 years 5 months
Refractory severe rejection
Refractory severe rejection
2 years 3 months
Refractory severe rejection
Surgical complications after pediatric LDLT using reduced and hyper-reduced grafts
Two (4.1%) patients suffered from significant HAT (Table 4). One died before the postoperative third week, and the other developed HAT on postoperative day (POD) 8 and was treated by revision of the anastomosis.
Table 4. Surgical complications after pediatric LDLT using reduced and hyper-reduced left lateral segment grafts
Number of cases (%)
Portal vein complications (thrombosis and stenosis)
Hepatic artery thrombosis
Hepatic vein stenosis
Biliary complications (fistula or stricture)
Eight patients (16.3%) suffered from significant PV problems (stenosis: 6, thrombosis: 2). Two patients suffered from PV thrombosis, one of whom was managed with re-exploration but died of sepsis on POD 28. The other patient developed late PV thrombosis, which was treated by surgical correction using an external iliac vein graft. Six patients suffered from PV stenosis: five were successfully managed by transhepatic balloon dilatation, one of the five died from sepsis and the sixth underwent retransplantation. The rate of the PV complication after April 2006 was 0.0% (none of seven cases), whereas that before April 2006 was 19.0% (eight cases of forty-two cases). However, the difference was not significant (p = 0.207). The WIT before April 2006 was significantly longer (before vs. after April 2006: 48.3 ± 14.1 min vs. 37.0 ± 2.8 min: p = 0.022); the modification of the PV anastomosis was relevant to shortening of WIT. There was no difference between patients with or without PV complication in GRWR (yes vs. no: 3.62 ± 0.79% vs. 3.42 ± 0.93%; p = 0.516) or donor body weight (yes vs. no: 54.3 ± 4.9 kg vs. 64.8 ± 2.1 kg; p = 0.052).
Two patients (4.1%) were diagnosed with hepatic vein stenosis, and both were successfully managed with transcaval balloon dilatation. Four patients (8.2%) experienced bile duct problems (fistula: 2, stricture: 2). The two patients with bile duct strictures both underwent successful percutaneous transhepatic biliary drainage and balloon dilatation. One of patients with a biliary leak required a re-exploration and revision of the bilio-enteric anastomosis, and the other responded to conservative management without any additional intervention required. Three patients (6.1%) required re-exploration due to bowel perforations, two of which were in the colon and one of which was in the small intestine. One patient developed graft versus host disease and improved on steroid pulse therapy. The incidence of early and late biopsy-proven ACR and chronic rejection in our study were 49%, 36.5% and 2%, respectively.
Prognostic factors for patient and graft loss after pediatric LDLT using reduced and hyper-reduced left lateral segmental grafts
Twelve items, including recipient, graft, operative and postoperative factors, were analyzed using both univariate (Table 5) and subsequently multivariate analysis (Table 6). Univariate analysis identified four potential prognostic factors: FHF as an indication for transplantation, a recipient age of less than 2 months, a body weight of less than 5 kg and a WIT of more than 40 min were each associated with poor patient survival. In the multivariate analysis with backward elimination, a recipient age of less than 2 months and a WIT of more than 40 min significantly affected patient and graft survival.
Table 5. Univariate analysis of risk factors for patient and graft survival after reduced and hyper-reduced liver transplantation
Table 6. Multivariate analysis of risk factors for patient and graft survival after reduced and hyper-reduced liver transplantation
Hazard ratio (95% CI)
WIT = warm ischemic time.
Recipient age, <2 months
WIT, >40 min
A major problem in pediatric LT has been the lack of size-matched donor organs; this shortage is most notable for the smallest children. Segmental liver reduction has increased the potential pool of grafts and has made a significant contribution to a reduction in the pediatric waiting list mortality rate. The use of a left lateral segment graft can overcome a donor-recipient size discrepancy of up to 10:1, but on occasion, even this option may not be sufficient for transplants in very young babies (5). The disadvantage of the use of large-for-size grafts is the insufficient tissue oxygenation and graft compression, which result in poor outcomes (10). To relieve the problem of large-for-size grafts in small babies, reduced grafts (partial left lateral segment, monosegment) or hyper-reduced grafts (reduced monosegment grafts) have been introduced. We described successful LDLT and long-term follow-up using these grafts and demonstrate potentially lower vascular complication rates.
We observed tolerable patient and graft survival rates in our series. The actuarial patient survival at 1, 5 and 10 years after LDLT was 83.7%, 78.9% and 78.9%, which was lower than pediatric patients in the same period (n = 246) with left lateral segment graft (patient survival at 1, 5 and 10 years after LDLT was 93.1%, 90.5% and 89.0%) (p = 0.033).
The patient survival after April 2006 in our center was better than previous rates, but this difference was not significant (p = 0.243).
In the case of reduced LT and split LT, however, several reports have cited technical difficulties in dealing with the small size of the graft vessels and the associated deterioration in outcomes, such as biliary complications and HAT, relative to complications associated with whole LT (2). LDLT is also considered to be technically challenging because partial liver grafts have small, short vessels that must be anastomosed (11). Early occurrence of HAT frequently leads to graft loss, with an incidence of 5–25% for pediatric patients. For LDLT, the small, short hepatic arteries in the partial liver with diameters of 2 mm or less are anastomosed, which may increase the risk of HAT. However, the introduction of microscopic anastomosis for HA reconstruction has resulted in a notably lower incidence of HAT (3.3%) (12). The incidence of HAT in our series was 4.1% (n = 2); one of these two patients died and the other was managed with a revision of the anastomosis. Stringer et al. (13) reported the incidence of HAT to be 7.8% in a study of 400 pediatric LTs, with 83% of these children being alive 3.6 years after the transplantation.
The incidence of PV complications in our series, either stenosis or thrombosis, was 16.3% (n = 8), higher than in our previous report of 600 pediatric LDLT patients (9%) (14) and lower than the 27% rate reported by Buell et al. (15) for pediatric LDLT. Long-term periodic examination with Doppler ultrasonography is mandatory for the early detection of perturbation of portal venous flow before completion of thrombotic occlusion, even if the children are asymptomatic with satisfactory liver function tests. We suspect that PV complications are high when using these grafts, due to the caliber discrepancy between the graft and native vessels, which may be further compounded by compression of the graft and reconstructed vessels. Another concern is the discrepancy in the distance from the portal vein to the hepatic vein, which is much longer in these grafts and may induce a kinking of the reconstructed portal vein, resulting in PV complications (4). Of note, we did not observe any PV complications after the modification of PV reconstruction from April 2006. We cut the recipient's PV diagonally and shortened it as much as possible to avoid twisting or bending. In addition, we ligated all collateral vessels to increase PV flow. We hypothesize that this approach is one of the factors contributing to increased survival. Long-term observation may be necessary to collect sufficient data for the establishment of such a modality.
The incidence of hepatic vein complications in our study was 4.1% (n = 2), which is nearly equal to that in our previous report of 600 pediatric LDLT patients (hepatic vein stenosis; 3.7%) (12), and both of these cases were successfully managed with radiological venoplasty.
The rate of biliary complications in our study was 8.2% (n = 4), which was comparable to that reported previously for pediatric LT using technical variant grafts with rates 5–30% (16). Only one of these patients required another surgery for revision of the hepato-jejunal anastomosis, indicating that the radiological treatments of biliary complications were markedly effective.
The incidence of biopsy-proven ACR (49%) in our series was as high as that in our previous study of 600 pediatric LDLT (52.8%) (4). Colombani et al. reported that infants younger than 1 year of age experienced a higher incidence of acute rejection (17). A recent study from Pittsburgh (18) revealed early ACR of liver allografts in children to be associated with enhanced donor-specific alloreactivity and to require additional immunosuppression.
In our study, we used segment III in all cases of monosegmental grafts. Srinivasan et al. (5) reported that segment III transplantation appears to be easier and safer from a technical standpoint because segment II transplantation requires extensive dissection at the base of the umbilical fissure to expose the portal venous and hepatic arterial inflow to segments II and III.
Ten patients died during the follow-up period. The most common cause of mortality was sepsis (40%) followed by rejection (30%). These findings are comparable to those reported by Jain et al. (19), with the most common cause of death in his study being infection. Two patients in our study died for reasons deemed unrelated to postoperative complications of LT.
Univariate analysis of prognostic factors for patient and graft survival revealed that FHF, age less than 2 months, body weight less than 5 kg and WIT > 40 min were associated with poor patient and graft survival. In the multivariate analysis, WIT > 40 min and age less than 2 month was the independent risk factor for poor patient and graft survival. No statistically significant differences were found for patient survival by donor relation, donor age, type of graft, GRWR, PELD, CIT, ABO incompatibility, presence of HAT, PV, biliary complications, HV complication, early ACR or the era at LT. Before introducing reduced or hyper-reduced grafts, we encountered a case in which GRWR was 9.68%. The graft could not obtain sufficient portal flow and massive hepatocyte necrosis occurred. This was the first patient we lost due to problems related to a large-for-size graft (2). However, GRWR in our study (2.26–5.29%) did not have an impact for the patient or graft survival, which shows the use of hyper-reduced graft with GRWR lower than 5% is acceptable for the small infant.
Patient survival after FHF was the lowest of that for all primary diseases. This lower patient survival for FHF is thought to be related to the high frequency of treatment-resistant rejection, and this phenomenon was particularly frequent among infants (12). Other reports have also highlighted that the major indication of LT for babies less than 3 months of age is neonatal hepatitis (20), and the incidence of corticosteroid-resistant rejection is known to be high among such patients (21). It remains unclear, however, why small infants with immature immunity and presenting with FHF demonstrate such severe rejection to donor grafts during immunosuppressive therapy. However, a definitive diagnosis of acute rejection and the administration of aggressive corticosteroid treatment to these patients seem to be effective. Although infantile FHF is one indication for LDLT in neonates, the results remain disappointing in many centers (22,23).
The rate of the PV complication after April 2006 (modification of the PV reconstruction) tends to be lower than that before April 2006 with significant shorter WIT. Further accumulation of cases and long-term follow-up are necessary to collect the necessary data to confirm these tendencies.
In summary, pediatric LDLT using reduced and hyper-reduced left lateral segment grafts appears to be a feasible option with acceptable rates of graft survival and vascular complication.
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.