Objectives: Hepatic venous congestion is associated with impaired graft regeneration in living-donor liver transplant, and the management of middle hepatic vein tributaries in the right lobe graft represents an unresolved issue. In this study, we aimed to investigate the precise outflow pattern of segments 5 and 8 between the right hepatic vein and middle hepatic vein and the respective regeneration rates after living-donor liver transplant with right lobe graft, as available data on these relevant topics are scarce.
Materials and Methods: We conducted a retrospective analysis of computed tomography scans with 3-dimensional simulation, vessel reconstruction, and volume measurement of 38 right lobe grafts without middle hepatic vein. Follow-up time was 3 months after living-donor liver transplant.
Results: In donors, segments 5 and 8 measured 141.9 ± 48.8 mL (21.0% of graft volume) and 230.4 ± 52.5 mL (34.3% of graft volume), respectively, with significant difference between volumes (P < .01). Percentage of segmental venous drainage in segment 5 was 55.5 ± 17.2% for the middle hepatic vein and 41.0 ± 20.9% for the right hepatic vein; drainage in segment 8 was 46.4 ± 13.2% for the middle hepatic vein and 52.9 ± 13.2% for the right hepatic vein. The outflow pattern was significantly different between segments for both veins (P = .01 for middle hepatic vein and P < .01 for right hepatic vein), showing that segment 5 was statistically more dependent on the middle hepatic vein and segment 8 was more dependent on the right hepatic vein. For living-donor liver transplant recipients, the prevalence of middle hepatic vein tributary reconstruction was 39.5%. At 3-month follow-up, the regeneration rate for the posterior sector was 85.8 ± 39.9%, whereas rates for segments 5 and 8 were 33.4 ± 39.7% and 68.4 ± 41.0%, respectively (P < .01).
Conclusions: In living-donor liver transplant with right lobe graft and without middle hepatic vein, segment 5 is the most vulnerable graft area for impaired regeneration. Segments 5 and 8 should be evaluated independently on the basis of their respective outflow patterns to more precisely plan the outflow management and patient outcomes.
Key words : Anterior sector, Hepatic venous congestion, Liver regeneration, Middle hepatic vein tributaries
In living-donor liver transplant (LDLT) procedures, the selection of the graft type represents a complex and strict process that must account for several variables to minimize the incidence of morbidity and mortality in both donors and recipients.1,2 Right lobe grafts (RLGs) have become a standardized choice to meet the metabolic demand of adult recipients due to the advantage of a large graft volume.1,3,4 However, the management of anterior sector (segment 5 and segment 8) outflow remains an unresolved issue.1-4 Available alternatives are inclusion of the middle hepatic vein (MHV) in the RLG, reconstruction of the MHV tributary branches for segment 5 and 8 (V5, V8), and no outflow restoration. The target is to minimize the grade of hepatic venous congestion in the graft while maintaining adequate safety for the living donor and having a cost-effective procedure.1-4
Hepatic venous congestion secondary to outflow obstruction exposes the parenchyma to both mechanical injury, due to acutely increased sinusoidal pressure, and ischemic injury, due to impaired portal inflow.5,6 As a result, metabolic dysfunction and impaired regenerative capacity may occur, exposing the recipient to liver insufficiency and septic complications in the early postoperative period.1-3,5 This damage is temporary thanks to the development of venous intrahepatic collaterals and is compensated by hyperregeneration of the noncongested area.2,3
Many studies have already investigated drainage volumes of V5 and V8, identifying them as expected congested areas when there is no reconstruction and thus as risk factors for impaired graft regeneration.6,7 Nevertheless, the precise drainage distribution of the right hepatic vein (RHV) and MHV when segments 5 and 8 are identified according to the portal inflow volume has not yet been analyzed. Moreover, so far the volumetric evolution of the graft has been examined in recipients just as a comparison between the volume changes of anterior vs posterior sector. However, segments 5 and 8 show complex and diversified outflow patterns that could predispose to different regeneration rates. In the present study, our aim was to investigate the drainage patterns and regeneration rates of segments 5 and 8 in a selected population of adult LDLT with RLG without MHV.
Materials and Methods
This was a retrospective study of prospectively collected data. From May 2006 to October 2016, 64 adult patients underwent LDLT with a RLG without MHV at the Nagasaki University Hospital. Exclusion criteria were patient or graft survival < 3 months after LDLT (4 patients excluded). Unavailability of a strictly 3-month follow-up contrast-enhanced computed tomography (CT) scan or in case of examination undertaken due to suspected surgical complications(22 patients excluded). The resulting study population included 38 patients, with patients divided into 2 study groups (Figure 1): (1) no reconstruction group, which included 24 recipients of a RLG without reconstruction of the MHV tributaries, and (2) reconstruction group, which included 14 recipients of a RLG with reconstruction of the MHV tributaries.Preoperative 3-dimensional computed tomography scan analysisImage analyses were performed using Synapse VINCENT (Fujifilm Medical, Tokyo, Japan) as previously reported.8,9 Retrospective preoperative and follow-up CT scans had been independently analyzed from previous simulations by a single researcher under the supervision of the surgeons who attended the LDLTs. A 3-month time interval was selected for follow-up imaging evaluation to minimize the effects of any possible confounding factors on volumetric measurements, such as hepatic venous congestion.
For reconstruction, 0.5- to 3-mm-thick images acquired during a 3-phase (arterial, portal, venous) dynamic multidetector CT were used. The stem of the main portal vein was set as a seeding point, and the direction was set for the peripheral side; the portal tree was then extracted automatically. To extract the peripheral thin branches for complete segmentation, a semiautomatic additional setting of seeding points was performed. Each hepatic vein was also extracted using the same procedure from the direction of the confluence to the inferior vena cava. After whole liver volume and vessel reconstruction, a virtual liver partition was performed following the course of the MHV as carving transection plane, exposing the vein at the resection surface. Thereafter, only the strict volume of the future graft was considered in donors. In both donors before LDLT and in recipients at 3-month follow-up after LDLT, the anterior sector was identified and measured as the perfusion volume of the right anterior portal branch. The graft area at the resection surface that was drained by the MHV but not perfused by the right portal vein was calculated and recorded as ischemic area. The posterior sector volume was calculated as subtraction from the total graft volume of the anterior sector and ischemic area. Thereafter, segments 5 and 8 were identified according to craniocaudal segmentation using the axial level plane of the portal main bifurcation as landmark (in recipients, this corresponded to the portal anastomosis). The segmental drainage volumes of the MHV, RHV, and inferior RHV were calculated by measuring the shared volume between the respective total volume drainages of the hepatic veins and volumes of segments 5 and 8.
Graft selection algorithm
In our department, the graft type selection algorithm is as follows: if the volume of extended left lobe is more than 30% of the recipient’s standard liver volume (SLV), then an extended left lobe graft is selected. Otherwise, a RLG would be the second option when the remnant liver volume is more than 30% of the whole liver volume.
Middle hepatic vein tributary reconstruction policy and technique
The criteria for reconstruction of V5 and V8 were based on volumetric data. The anticipated congested volumes of these territories were calculated and subtracted from the total graft volume. The estimated noncongested graft volume was expected to be always greater than 40% of the SLV. Therefore, reconstruction of V5 and/or V8 was planned to achieve at least 40% of the SLV. Moreover, if the absolute drainage volume of these branches was more than 100 mL, reconstruction was also undertaken.
Three techniques were alternatively used for reconstruction of the V5 and V8 anastomosis with recipient’s MHV, direct anastomosis to inferior vena cava with graft patch venoplasty, or interposition venoplasty using autologous veins (portal vein or umbilical vein) or artificial graft. These techniques have been previously described.10-12 A diagnosis of hepatocellular carcinoma excluded the possibility of using the recipient’s portal veins or MHV. Otherwise, technique selection was based on intraoperative anatomic findings.
Continuous variables are expressed as means and standard deviation or as medians and interquartile range (IQR). Regeneration rates were calculated by comparing whole and segmental graft volumes at 3 months after LDLT, using the following formula: regeneration rate = ([3-month volume – preoperative volume]/preoperative volume) × 100.
Statistical analyses were performed with the SPSS software program (IBM SPSS Inc., Chicago, IL, USA). For categorical variables, cross-tabulations were generated, and chi-square or Fisher exact test was used to compare distributions. For continuous variables Student's t test or Mann-Whitney test were used. P < .05 was considered significant.
Table 1 shows donor and recipient demographic and clinical characteristics. There were no significant differences between the 2 study groups for the variables analyzed. Overall, male-to-female ratio in the recipient population was 30:8 with a mean age of 54.4 ± 10.4 years. The median Model for End-Stage Liver Disease score was 15.5 (IQR, 12-19), and the prevalence of hepatocellular carcinoma diagnosis was 52.6%. Hepatitis C and hepatitis B virus infections were the causes of underlying liver cirrhoses in 31.6% and 26.3% of patients, respectively. The mean operative time was 813.0 ± 145.4 minutes, and the median length of hospital stay was 47 days (IQR, 38-55 d).
Graft volume and drainage patterns
The mean graft volume in the total population was 673.8 ± 115.1 mL, which corresponded to 53.8 ± 9.2% of the recipient’s SLV. No significant differences were shown between graft volume in the no reconstruction and the reconstruction groups (P = .85). The anterior sector volume was 372.4 ± 83.3 mL (55.2% of graft volume). The ischemic area was 14.8 ± 10.4 mL. Segments 5 and 8 showed volumes of 141.9 ± 48.8 mL (21% of graft volume) and 230.4 ± 52.5 mL (34.3% of graft volume), respectively, with a significant difference between volumes (P < .01) (Figure 2).
Single or multiple inferior RHVs were present in 68.4% of grafts, but the median percent drainage volume in the anterior sector was 0 (IQR, 0-1.2%); thus, this measurement was not included in the analyses. Segmental venous drainage patterns (see Table 2) were as follows: 55.5 ± 17.2% for MHV and 41.1 ± 20.7% for RHV in segment 5 and 46.4 ± 13.2% for MHV and 52.9 ± 13.2% for RHV in segment 8. These percentage drainage volumes were significantly different for both MHV and RHV between the 2 segments (P = .01 for MHV and P < .01 for RHV), showing that segment 5 was statistically more dependent on MHV and segment 8 was statistically more dependent on RHV (Figure 2).
Comparisons of MHV percentage drainage volumes between the 2 groups showed that the total and segment 8 percentage drainage volumes in the reconstruction group were significantly greater than in the no reconstruction group (P < .01 for percentage drainage volume and P < .01 for segment 8 percentage drainage volume). The MHV percentage drainage volume in segment 5 was similar between the 2 groups (P = .32). This result was further confirmed by a higher prevalence of V8 vs V5 reconstruction at transplant (Figure 1). Occlusion of V8 occurred in 60% of reconstructed vessels at a median time of postoperative day 4 (IQR, 2-6 d). The patency of V5 was 100% in the evaluated time interval.
At 3 months after LDLT, the mean graft volume was 1092.8 ± 182.9 mL, with posterior sector volume of 525.0 ± 133.7 mL, segment 5 volume of 183.3 ± 70.6 mL, and segment 8 volume of 384.5 ± 117.8 mL. The overall and group-specific regeneration rates are shown in Table 3 and Figure 3.
In both the total population and within the specific study groups, the posterior sector showed a significantly higher regeneration rate versus the anterior sector. In the no reconstruction group, the regeneration rate of segment 5 was statistically lower than the rate of segment 8, with some cases of even segmental atrophy (negative regeneration rate). In the reconstruction group, differences in regeneration rates between the anterior and posterior sectors were slightly significant (P = .04). Moreover, the regeneration rates of segments 5 and 8 appeared to be balanced by preoperative selection criteria and reconstruction of V5, V8 reconstruction and showed no significant differences. Among cases with reconstruction of V8, the occurrence of vein occlusion was associated with a lower regeneration rate of segment 8 (6 patients, 60.1 ± 34.3%) compared with those with patent vein at 3-month follow-up (4 patients, 81.4 ± 54.7%). However, these differences were not statistically significant (P = .52), possibly due to the small number of cases.
When we compared the reconstruction versus the no reconstruction groups, we observed no significant differences between either whole graft or segmental volumes in terms of regeneration.
Graft regeneration after LDLT is a complex process that is modulated by multiple factors such ischemic injury, graft size, immunosuppression, steatosis, donor age, and viral hepatitis.13 Hepatic venous congestion also represents a risk factor for impaired regeneration, and its relevance is related to its direct clinical correlation with the surgical technique of outflow reconstruction.1,2,4,5,13
Retrieval of the MHV for a RLG guarantees the best outcome for the recipient but may expose the donor to postoperative liver failure.1-3,7 Reconstruction of the segment 5 and 8 MHV tributaries is an effective but time-consuming procedure at the back-table, which may prolong graft ischemia time and may not guarantee a postoperative long-lasting patency.2,4,6,7 In our series of patients, the patency rate of reconstructed V5 and V8 at 3 months after LDLT was 100% and 40%, respectively (60% if evaluated together as MHV tributaries). Akamatsu and associates,7 in a series of 126 RLG LDLT procedures with MHV tributary reconstruction using cryopreserved homologous veins, recorded a patency rate over the same time frame of 69% and 77% for V5 and V8, respectively. Hwang and associates12 reported the results of MHV reconstruction with various graft types, including cryopreserved iliac veins, iliac arteries and aortas, and polytetrafluoroethylene grafts, in 262 cases. The group reported 6-month patency rates of 75% in the iliac vein group, 35% in the iliac artery group, 92% in the aorta group, and 77% in the polytetrafluoroethylene group.
In evaluations of hepatic venous congestion as a risk factor for impaired regeneration, there are many additional variables, including ischemic dysfunction, patency of the reconstructed vessels, timing of occlusion, and compensatory hyperregeneration of the noncongested areas. In the MHV reconstruction group of the present series, the prevalence of V8 reconstruction was correctly higher than V5 based on the MHV percentage drainage volume to segment 8 and segment 5. However, its patency rate was already dismal early postoperatively. Conversely, although fewer in number, the reconstructed segment 5 tributaries remained patent. In the MHV reconstruction group, segment 5 regenerated with a trend that was better (although not statistically significant) than segment 5 in the no MHV reconstruction group, despite a similar percentage drainage volume. Unfortunately, the limited number of cases in our series did not allow us to determine whether this trend was due to a correct preoperative selection, high patency prevalence, or compensation to a congested segment 8. Despite this limitation, we believe that our results still add some new insight on the mechanisms of segmental regeneration, with furtherinvestigations needed on larger study populations.
There is substantial agreement on the criteria for MHV procurement but not on segment 5 and 8 MHV tributary reconstruction. Many institutional policies have been published to establish the optimal outflow selection process variably using as parameters the hepatic vein dominance, graft-to-recipient weight ratio, remnant liver volume, donor-to-recipient body weight ratio, volume of the donor’s right lobe-to-recipient’s SLV ratio, size of MHV tributaries, and expected congested volume-to-recipient SLV ratio.1,3,5 So far, no consensus has yet been achieved,1,2,4,7 and impaired regeneration rates due to hepatic venous congestion are still reported in LDLT recipients selected for RLGs without MHV or reconstructed segment 5 and 8 MHV tributaries.1,2,6
The efficacy of our policy for MHV tributary management was validated even when meticulous vessel reconstruction and identification of segment 5 and segment 8 according to portal inflow volumes were performed. The policy succeeded to select and thereafter reconstruct the MHV tributaries of grafts with segment 8 that were more dependent on MHV. The same effectiveness was not reached for segment 5 as the MHV percent drainage volumes in this segment were comparable between the groups with and without reconstruction. Nevertheless, in the reconstructed group, the MHV percentage drainage volumes were similar in segments 5 and 8 and the segmental regeneration rates were not statistically different. Furthermore, at 3 months, the total and segmental volumes of the graft were comparable between the reconstruction and the no reconstruction groups; therefore, the reconstruction policy was effective. These results on whole graft outcomes are in line with previous reports.1,7
In conclusion, hepatic venous congestion is a risk factor for segmental impaired regeneration in LDLT. The midterm outcomes for the whole graft appear not to be affected, probably due to the development of intrahepatic collateral vessels and compensatory hyperregeneration of the noncongested graft areas. In RLG without MHV, segment 5 is the most vulnerable graft area for impaired regeneration. The related pathogenesis is probably multifactorial, but an outflow pattern more dependent on MHV than RHV may be a possible mechanism. Thus, we believe that segment 5 and segment 8, both preoperatively and after LDLT, should be analyzed independently to more precisely evaluate the outcome. Moreover, the reciprocal drainage distribution of MHV and RHV in segments 5 and 8, when identified according to the portal inflow volume, may be a more effective method to select tailored indications for V5 and V8 reconstruction.
DOI : 10.6002/ect.2018.0155
From the 1Department of Surgery, Nagasaki University Graduate School of
Biomedical Sciences, Nagasaki, Japan; and the 2General Surgery and
Transplantation Unit, Department of Medicine, University of Udine, Udine, Italy
Acknowledgements: The authors have no sources of funding for this study and have no conflicts of interest to declare. The authors thank Giada Aizza for her support.
Corresponding author: Susumu Eguchi, Department of Surgery, Nagasaki University Graduate School of Biomedical Sciences, 1-7-1 Sakamoto, Nagasaki 852-8501, Japan
Phone: +81 95 819 7316
Figure 1. Reconstruction and No Reconstruction Groups Showing Prevalence of Middle Hepatic Vein Segment 5 and 8 Tributary Reconstruction and Patency Rate at 3 Months Posttransplant
Figure 2. Graft Volume in Segments 5 and 8
Figure 3. Graft Segmental Regeneration in the Study Groups
Table 1. Demographic, Clinical, and Operative Characteristics of Donors and Recipients
Table 2. Comparisons of Venous Drainage Patterns in the Total Population and Study Groups
Table 3. Comparisons of Whole Graft and Segmental Regeneration Rates in the Total Population and Study Groups