Address reprint requests to Juan Carlos García-Valdecasas, M.D., Ph.D., Hepatobiliary Surgery and Liver Transplantation Unit, Hospital Clinic of Barcelona, Villarroel 170, Barcelona, Spain 08036. Telephone: +34932275718; FAX: +34932275589; E-mail: firstname.lastname@example.org
Like all types of partial graft transplantation, adult living donor liver transplantation (aLDLT) includes the risk of developing so-called small-for-size syndrome (SFSS), which is characterized by a progressive impairment of the synthesizing capacity of the liver and progressive liver failure. This possibility is inversely proportional to the graft size. An excessively small graft is incapable of meeting the metabolic demands of the recipient; thus, it is important in a clinical setting to ensure the correct relationship between the patient's weight and the graft size [graft-to-body weight ratio (GBWR)] as well as the hemodynamic condition of the recipient. Nowadays, the consensus is that grafts are small for size when the GBWR is less than 0.8% or the standard liver volume is less than 35%. Furthermore, a number of studies have shown that these grafts are related to poor survival rates.
The safety of the donor is undoubtedly one of the most important and controversial aspects of aLDLT. The morbidity rate associated with donation of the right liver lobe has been estimated to be 27% to 67%, and up to 38% of these morbidities are Clavien-Dindo classification stage I or II. Recent studies by Iida et al. and Belghiti et al. have confirmed that the most important factor in the appearance of complications is the magnitude of the hepatectomy (44.2% of overall complications in right lobe donors vs. 18.8% in left lobe donors, P < 0.05). The possibility of reducing the surgical risk in the donor necessarily involves reducing the size of this surgical operation.
It is obvious that if the left liver (representing 35% of the total volume) could guarantee the viability of the patient despite its small size, this would considerably reduce the size of the operation and thus the objective risk for the donor. However, to date, the use of the left lobe in adults has been limited because of its frequent association with SFSS.
On the other hand, the recipient of a liver transplant is generally a patient with advanced cirrhosis. The presence of portal hypertension in these patients is characterized by a hyperdynamic state with high cardiac output and reduced peripheral vascular resistance. After the native liver hepatectomy, the hyperdynamic state is temporarily maintained, so a high portal vein flow (PVF) persists. The combination of these 2 factors—a partial (small) graft and a high PVF—may lead to a situation of excessively high flow with respect to a reduced vascular bed and thus increase the risk of SFSS.[4, 13]
From a physiopathological point of view, these facts are of prime importance, and so transplant groups working in Asia have attempted to control the excess flow after reperfusion. These groups have been able to demonstrate that the portal pressure (PP), an indirect indicator of the flow, is a survival factor to the extent that at present, they suggest that it should be approximately 15 mm Hg at the end of the operation. Furthermore, a Hong Kong group has established that as long as PP is within this range, the graft size can be considerably reduced to a GBWR of 0.6%, and this makes it possible to use the left lobe of the donor with the same level of safety.
On the other hand, different authors have suggested that measuring only PP may not be adequate because a low PP does not necessarily mean an appropriate PVF. Instead, the portal pressure gradient (PPG) or the hepatic venous pressure gradient (HVPG) in addition to PVF may be better for evaluating graft compliance to hemodynamic stress, and this has been confirmed in both clinical and experimental studies.
Our group has shown in different experimental works that several factors must be taken into consideration. In the first place, when we are dealing with a partial graft, there is a need for a relative increase in the PVF rate, and this is known to be related to survival.[18, 20] However, if this flow is excessive, it might immediately damage the vascular endothelium and, depending on the degree, might lead to the development of SFSS. Two important conclusions can be derived from these experimental results: PVF at the time of reperfusion is of paramount importance, and it would be most interesting to know this information actually before reperfusion in order to be able to prevent excessive portal blood flow and subsequent endothelial injury.
From the beginning of our aLDLT program, we have systematically performed hemodynamic monitoring of our patients during the surgical procedure. Our purpose has been to control all factors that can place the graft in the most favorable situation at the time of reperfusion (appropriate PVF and artery flow). This surgical protocol includes the systematic performance of a portocaval shunt during the anhepatic phase. Recently, the possibility of modifying these conditions through what is called graft inflow modulation has given greater importance to hemodynamic monitoring during the transplant procedure.
The purposes of this study were to analyze hemodynamic parameters during liver transplantation with a living donor graft and to validate their usefulness for deciding what measures to take to improve posttransplant outcomes. In this sense, the hemodynamic information obtained with a portocaval shunt during the anhepatic phase of the procedure may be of great interest.
PATIENTS AND METHODS
The aLDLT program and all the clinical protocols derived from it were approved by the hospital's ethics committee at the start of the program. Forty-five consecutive patients undergoing right lobe aLDLT from 2003 to 2011 were analyzed. No patient was excluded, and these patients represent the full series of aLDLT procedures performed at our center during that period.
The data were collected prospectively and were analyzed retrospectively. Table 1 shows the demographic characteristics as well as the indications for transplantation.
Table 1. Demographic Characteristics of Recipients of Living Donor Liver Transplantation
aThe data are presented as means and standard deviations.
The surgical procedure for the donor was previously published. Briefly, it consisted of a right hepatectomy (liver segments V-VIII) with preservation of the middle hepatic vein, which remained in the donor. After an ultrasonographic and cholangiographic study, the elements of the hepatic hilum were identified, and the right hepatic vein was isolated. Liver parenchymal transection with a Cavitron ultrasonic surgical aspirator (Tyco Healthcare, Mansfield, MA) and TissueLink (TissueLink Medical, Inc., Dover, NH) was performed. Great care was taken to identify the branches of the middle hepatic vein with a diameter of more than the 5 mm. All were revascularized in the recipient with cryopreserved grafts of the iliac artery to achieve the best possible venous drainage. Once the graft was extracted, it was perfused with Celsior solution through the portal vein.
The surgical operation for the recipient consisted of 2 well-differentiated stages: hepatectomy of the native liver and graft implantation.
Before the hepatectomy, a 16-G venous catheter with a length of 30 cm (Certofix Mono S 330, Braun) was placed into the mesenteric vein, and the tip was advanced into the portal vein. This catheter was used to measure PP during the operation.
Hepatectomy was performed as previously published with systematic preservation of the inferior vena cava as well as a temporary terminolateral portocaval shunt, which was routinely sectioned before graft implantation.
The implantation began with a wide (4-cm) end-to-lateral anastomosis between the right hepatic vein of the graft and the recipient's vena cava. If it was necessary, anastomosis of the accessory veins was performed independently of the vena cava. This was followed by sectioning of the portocaval shunt and the performance of an anastomosis between the portal vein of the recipient and the right portal vein of the graft. The portocaval shunt was left in no patients after graft implantation, and none of the patients required ligation of preexisting spontaneous portocaval shunts. Finally, the hepatic artery anastomosis and the bile duct anastomosis were completed.
Hemodynamic monitoring of liver transplant recipients was undertaken only for those receiving living donor grafts at our center. No complications were experienced by the recipients as a result of their compliance with the protocol. The protocol consisted of 2 differentiated stages.
First, the recipient's hemodynamic measurements were taken by the hepatic hemodynamics laboratory 24 hours before the surgical operation, on the third day after transplantation, and in the third month. The aim of these evaluations was to provide extensive knowledge of the hemodynamic status of the recipient before surgery and to ascertain the normalization of the patient's portal hypertension after liver transplantation.
Second, an intraoperative hemodynamic assessment was performed. Hemodynamic measurements were taken in 5 well-defined phases: at the baseline, during the anhepatic phase, after portal reperfusion, after hepatic artery reperfusion, and at the end of the operation. All collected hemodynamic parameters are shown in Table 2. Systemic hemodynamic parameters were continuously monitored with invasive methods (an arterial line and a pulmonary catheter). After the vascular anastomoses were completed, hemodynamic measurements were taken for all patients after an arbitrarily defined 5-minute period.
Table 2. Variables Analyzed for the Hemodynamic Monitoring Protocol
Pulmonary artery pressure
Wedged hepatic venous pressure
Pulmonary capillary wedge pressure
Free hepatic venous pressure
Right atrial pressure
Median arterial pressure
Systemic vascular resistance
Systemic vascular resistance index
Pulmonary vascular resistance
Pulmonary vascular resistance index
Time of Measurement
Onset of the intervention
Portal vein pressure
After portocaval shunt
After portal reperfusion
Hepatic artery flow
After arterial reperfusion
After splenic artery ligation
Median arterial pressure
End of the operation
Central venous pressure
Wedged hepatic venous pressure
Free hepatic venous pressure
Flow rate measurements were made with the method for determining the transit time flow, with 8-mm to 12-mm probes used for the portal vein and 3-mm to 5-mm probes used for the hepatic artery; they were monitored with the VeriQ 1001 system (Medi-Stim ASA, Oslo, Norway).
The purpose was to achieve a PVF less than 2000 mL/minute at the time of reperfusion and (as long as the weight of the graft was known) a relative PVF less than 300 mL/minute/100 g of liver tissue. An arterial flow less than 100 mL/minute was arbitrarily considered insufficient.
Splenic artery ligation was performed in 13 patients for excessive PVF (n = 4), inadequate arterial flow (n = 3), or a combination of the two (n = 6).
HVPG was measured preoperatively and postoperatively on the third day and in the third month at the hemodynamics laboratory. It was defined as the hepatic wedged pressure (equivalent to PP) minus the hepatic free pressure (equivalent to the central venous pressure).
PP was measured continuously with a Philips IntelliVue MP70 anesthesia monitor (Philips MedizinSystems, Böblingen, Germany) through a portal vein catheter placed at the beginning of the intervention. The normal range for directly measured PP values was considered to be 7 to 12 mm Hg.
PPG was defined as the PP directly measured during the surgical intervention minus the central venous pressure.
Analytical determinations of the liver function [aspartate aminotransferase, alanine aminotransferase (ALT), alkaline phosphatase, gamma-glutamyltransferase, total bilirubin, and prothrombin index] were made in the immediate postoperative period on days 1, 2, 3, 5, 7, and 10 and later in the first and third months after surgery. Patients remained in the intensive care unit under continuous observation during the first few days. For the purposes of this study, a follow-up period of 3 months after transplantation was taken into consideration.
The diagnosis of SFSS used the clinical and analytical criteria described by Dahm et al. (Table 3).
Table 3. Criteria for SFSS
aDysfunction is defined as the presence of 2 of the following criteria on 3 consecutive days: bilirubin level > 100 μmol/L, international normalized ratio > 2, and grade 3 or 4 encephalopathy.
bThe exclusion criteria include technical criteria (eg, arterial or portal occlusion, outflow congestion, or bile leak), immunological criteria (eg, rejection), and infections (eg, cholangitis or sepsis).
cFailure is defined as retransplantation or death.
NOTE: This table was adapted with permission from American Journal of Transplantation. Copyright 2005, John Wiley & Sons, Inc.
Dysfunctiona of a small graft (GBWR < 0.8%) during the first postoperative week after the exclusion of other causesb
Failurec of a small graft (GBWR < 0.8%) during the first postoperative week after the exclusion of other causesb
The statistical differences between groups were analyzed with the Student t test for numerical variables and with Pearson's χ2 test for categorical variables. The statistical correlations between continuous variables were analyzed with the Pearson correlation test. All the results are expressed as means and standard deviations unless otherwise specified. A measurement was considered statistically significant if P was <0.05. The statistical analysis of the data was performed with SPSS Statistics 17.0 (SPSS, Inc., Chicago, IL).
No patient died during the immediate postoperative period. Progress during the first few weeks was satisfactory without any complications associated with the surgical procedure or immediately altered liver function in terms of severe dysfunction or rejection. The mean length of stay for the patients was 32.4 ± 18.9 days.
Two patients died during follow-up: one died 6 weeks after the transplant because of acute severe hepatitis of an unknown cause, and the other died 3 months after the transplant as a result of infectious complications (but the graft was functioning).
The GBWR was 1.04% ± 0.23% with a median value of 1.03% (range = 0.66%-1.83%). Four patients had a GBWR less than 0.8% (0.66%, 0.68%, 0.69%, and 0.79%). The median graft weight was 740 g (range = 507-1480 g).
Intraoperatively, 9 of the 45 patients required additional venous drainage reconstruction to ensure the best possible outflow: in 5 patients, an additional anastomosis between the vena cava and an accessory caudal hepatic vein was performed, and in 4 patients, an anastomosis between the vena cava of the recipient and a vein-draining segments V or VIII to the middle hepatic vein was performed using a cryopreserved iliac artery graft.
None of the patients fulfilled the clinical criteria for SFSS. All patients included in the study presented with an uneventful postoperative course and liver function within the usual parameters.
The immediate postoperative liver function in terms of the ALT and total bilirubin levels and the prothrombin index is shown in Fig. 1. As expected, except for total bilirubin, the variables showed a trend toward normalization after an initial period of 24 to 48 hours of deterioration.
Intraoperative Hemodynamic Assessment
Figure 2 shows the progress of the PVF rate during the operation. It is interesting to note that even though the recipients initially presented a baseline PVF rate of 727 ± 596 mL/minute, the hemodynamic situation changed immediately after the total hepatectomy of the native liver and the performance of the temporary portocaval shunt.
As can be seen in Fig. 2, after the temporary portocaval shunt was performed during the anhepatic phase, there was a significant increase in the absolute PVF and the relative PVF (with respect to the graft weight): they rose significatively in comparison with the baseline values (727 ± 596 versus 1362.8 ± 410 mL/minute, P = 0.01 for the absolute values and 69 ± 62 versus 204.2 ± 72 mL/minute/100 g for the relative values) and remained there until the end of the procedure. The values during 3 different time periods (during portal reperfusion, during arterial reperfusion, and at the end of transplantation) remained similar.
PP and PPG
Figure 3 shows the progression of PP and PPG during the operation. All patients showed a significantly increased PP at the beginning of the procedure (27.2 ± 4.2 mm Hg). Just like the PVF rate, after the temporary portocaval shunt was performed, PP dropped significantly (27.2 ± 4.2 versus 17.4 ± 6.1 mm, P < 0.001), and so did PPG (16.5 ± 5.9 versus 8.8 ± 5.3 mm Hg, P < 0.001). Moreover, the PP and PPG values showed a trend of dropping further after reperfusion of the graft (although the trend was not significant), with values of 14.9 ± 5.6 and 7.4 ± 4.4 mm Hg, respectively. The mean PP and PPG values at the end of the operation were 15 ± 4.6 and 7.9 ± 3.9 mm Hg, respectively. Despite adequate portal and arterial flow measurements, 4 patients presented with portal vein PP greater than 20 mm Hg at the end of the operation, and the PPG values were greater than 15 mm Hg.
Factors Related to PVF During Reperfusion
In order to identify patients at risk for an excessive PVF rate during reperfusion, we tried to find a correlation between the hemodynamic factors before reperfusion and PVF during graft revascularization. Among the hemodynamic factors studied, statistically significant correlations were found for the following:
PP of the Native Liver
A correlation between the PVF after reperfusion and the PP of the native liver was found (Fig. 4). This correlation was significant for the absolute PVF (R2 = 0.21, P = 0.007) and for the relative flow per 100 g of tissue (R2 = 0.22, P = 0.005).
PVF Rate With the Temporary Portocaval Shunt
Perhaps the most important observation of this study is the direct correlation found between the PVF at the temporary portocaval shunt and the portal blood flow at the time of reperfusion (R2 = 0.3, P < 0.001). Likewise, the flow obtained in the portocaval shunt during the anhepatic phase was related to the PVF per 100 g of tissue during reperfusion (R2 = 0.3, P = 0.001; Fig. 5).
Hemodynamic Monitoring After Transplantation
The hemodynamic assessment after transplantation showed a progressive return to normality for hepatic hemodynamic parameters. As soon as 3 days after transplantation, most patients experienced a significant decrease in portal hypertension in comparison with preoperative values (18.0 ± 4.3 versus 7.6 ± 2.9 mm Hg). At 3 months, 76% of the patients had an HVPG within normal values (<5 mm Hg); the mean was 4.4 ± 2.5 mm Hg (range = 1.5-10 mm Hg; Fig. 6).
aLDLT has 2 major controversial issues. On the donor side, there is an objective surgical risk, which is in direct relationship to the magnitude of the operation. Right hepatectomy in itself involves morbidity and mortality rates that are considerably higher than those associated with left hepatectomy. On the recipient side, the appearance of SFSS is one of the most feared complications with a partial graft. Although the relatively small size of the graft is the major factor, other factors, such as a prolonged ischemia time, the presence of steatosis (which can be avoided in living donor liver transplantation), and the poor clinical condition of the recipient, may also play an important role in certain clinical situations.[19, 25]
The physiopathology of SFSS is complex; however, it is known that one of the most important factors is excessive flow at the moment of reperfusion, which results in damage to the sinusoidal vascular endothelium and ischemia of the hepatocyte cells (which, if it is severe enough, leads to functional failure of the graft). Thus, it is important to know the hemodynamic situation before reperfusion.
A pretransplant hemodynamic evaluation and hemodynamic monitoring during the procedure may facilitate important information not only about the degree of portal hypertension but also about the vascular status of the splanchnic area. This type of information may be of paramount importance if we are to avoid excessive portal blood flow after reperfusion.
Although the hemodynamic situation may be suggested by the overall clinical situation (including measurements such as the cardiac index and the peripheral vascular resistance index), to date there is no reliable factor (either preoperative or perioperative) that can be used to precisely identify the splanchnic hemodynamic situation and allow an adequate surgical approach, especially when a partial graft is being used. Current clinical experience is limited to altering the hemodynamic conditions once the graft is reperfused. The data in this study establish a direct relationship between the flow in the portocaval shunt and the flow during reperfusion of the partial graft, and they suggest for the first time that performing this temporary portocaval shunt during the anhepatic phase enables us to know the conditions before reperfusion, and determine whether they will be favorable or not (an appropriate PVF), and thus alter them to prevent any injury to the vascular endothelium (which appears, as we have shown previously, immediately after reperfusion). The experimental model shows that guaranteeing the recovery of the graft requires a relative increase in PVF. However, if this is excessive, it may be detrimental. Our group has been able to demonstrate in an experimental model of SFSS in pigs that this increase should be 2 to 3 times the initial PVF rate measured in the donor. Other groups have also confirmed this in clinical practice. Although it is still a matter of debate, the relative PVF considered to be the most suitable is approximately 250 to 300 mL/minute/100 g of liver tissue with a maximum value of 360 mL/minute/100 g of liver tissue.
To date, groups with greater experience have given considerable importance to the PP at the end of the operation. Early in their experience, they showed that a pressure greater than 20 mm Hg at the end of the operation was related to lower survival. Recently, these same authors demonstrated that a PP of 15 mm Hg was also associated with improved survival. They recommended achieving these values via splenectomy. This study is important because it establishes that when PP is within this range (<15 mm Hg), the size of the graft can be considerably reduced to a GBWR of approximately 0.6%.
This opens up the possibility of systematically using smaller grafts. Some groups have shown that the use of left grafts in adults may be safe despite portal hyperperfusion and a lack of graft inflow modulation; however, the lack of good systematic results in the literature means that most groups (including those in Asian countries) have relied on right lobes in order to guarantee adequate outcomes. The Fukuoka group recently showed that right living donor liver transplantation and left lobe living donor liver transplantation had comparable outcomes as long as the graft volume/standard liver volume ratio was greater than 35%. Despite their good results, the authors recognized that they had been using some type of graft inflow modulation (splenic artery ligation, splenectomy, or a permanent hemiportocaval shunt) and that criteria for performing such graft inflow modulation are lacking. In the United States, some other authors recently published their experience with the use of left lobes in adults, and they altered the flow by performing a partial portocaval shunt. Unfortunately, they also relied on PP, and the portal blood flow was measured in only 4 patients. Furthermore, the need for late occlusion and reoperation of the shunt due to the appearance of encephalopathy caused by excessive shunting or insufficient PVF has been described for this type of patient.
The common factor for all these groups (apart from their modifications relying on PP) is the fact that readings were taken once the graft had been reperfused, that is, when the endothelial damage may already have occurred. Our group and others believe that a normal or almost normal PP does not necessarily mean a suitable PVF; for this reason, we believe that it would be better to monitor PVF along with PP or PPG, preferably before reperfusion, in order to be able to modify it if that is necessary.
A recent clinical study by Sainz-Barriga et al. showed that the PP and flow rate do not always have a significant correlation, and the PP does not adequately reflect the hemodynamic situation in the graft. These authors described specific cases in which there was excessive PVF with a normal PP. They also pointed out that up to 25% of patients with an excessive PP (>20 mm Hg) had low PVF rates (<90 mL/minute/100 g). In conclusion, as these authors suggested previously, we think that both measurements, on which all modifications should be based, should be taken into account.
PP and especially PPG can undoubtedly be used to measure the compliance of the graft. However, a reduction in the compliance could be due to 2 factors: excessive incoming blood flow or inadequate venous drainage. Even though all groups are seeking the best possible drainage, this may be relative and sometimes even insufficient. This further stresses the importance of knowing the PVF rate during the anhepatic phase because this reading is made in the absence of any resistance. On the basis of the relationship found in this study, this information should be used to determine that there is no problem with the outflow drainage. The objective is to maintain, together with a normal or nearly normal PP or PPG, a flow similar (if it is satisfactory) to that obtained during the anhepatic phase in the absence of any type of modulation. In our experience, PVF not only is maintained but also tends to increase after reperfusion and, at the same time, reduce PP. The hemodynamic study on the third postoperative day allowed us to evaluate this trend to normality (7.5 mm Hg), which was later confirmed at 3 months.
In summary, hemodynamic monitoring during liver transplantation with a partial graft is necessary. In this context, we believe that a portocaval shunt should be performed systematically during the anhepatic phase. Knowledge of the PVF rate during this phase enables a precise determination of the conditions at the moment of reperfusion and thus allows any type of hemodynamic alteration to be made before reperfusion to ensure that the graft is in the most favorable situation. Moreover, this measured information enables us to ensure the best possible outflow drainage.