Objectives: Small-sized grafts are associated with high rates of graft failure and small-for-size syndrome. Portal flow is a causative factor for small-for-size syndrome. We sought to evaluate early graft dysfunction in smaller-sized grafts and the study factors responsible for it.
Materials and Methods: A total of 450 patients underwent a living-donor liver transplant from January 2010 to June 2013. Fifty-four grafts with graft/recipient’s body weight ratio less than 0.8 were included in the study. We used a splenic artery ligation or splenectomy for portal flow modulation if the portal flow after reperfusion was greater than 250 mL/min/100 g. Small-for-size syndrome was defined according to Clavien and Kyushu university definitions. Portal flow was measured with Doppler ultrasound flowmetry. Factors responsible for early graft dysfunction also were analyzed.
Results: Six patients out of 54 developed small-for-size syndrome in smaller size group (graft/recipient’s body weight ratio < 0.8). There were 28 left lobe grafts and 26 right lobe grafts. Sixteen out of 132 patients from the control group fulfilled the definitions of small-for-size syndrome. There was no statistical significant difference in graft dysfunction between low graft/recipient’s body weight ratio group and high graft/recipient’s body weight ratio group. On univariate analysis Hepatitis C, Hepatitis B and HCC as etiologies, Model for End-stage Liver diease score, and portal flow achieved statistical significane as factors associated with graft dysfunction (P < .05). On multivariate analysis, only portal flow achieved statistical significance.
Conclusions: Lower graft/recipient’s body weight ratio graft with portal flow modulation in case of high portal flow is an effective way to increase donor pool and donor safety with low risk of small-for-size syndrome. Portal flow is mainly responsible for small-for-size syndrome or early graft dysfunction.
Key words : Portal hemodynamics, Living donor, Small graft
There is increasing evidence that pathophysiology for small-for-size syndrome depends on portal hemodynamics rather than graft size.1,2 Troisi and associates2 report good results if portal flow was modulated to less than 250 mL/min/100 g. Kaido and associates suggest that crafts up to 0.6 graft/recipient’s body weight ratio (GRWR) with portal pressure control could be used with acceptable recipient survival rates and donor complication rates.3 We examined the effect of using a small-for-size graft with portal inflow modulation when portal flow was greater than 250 mL/min/100 g as described by Troisi and associates2 and its effect on developing small-for-size syndrome as per standard definitions.
The aims of the study were to evaluate the occurrence of small-for-size syndrome or early graft dysfunction as per definitions after portal flow modulation when portal flow was greater than 250 mL/min/100 g. We also tried to study factors responsible for early graft dysfunction regardless of GRWR status by comparing early graft dysfunction in small-for-size group with the control group, and also tried to evaluate accuracy of the cutoff value of portal flow for portal flow modulation.
Materials and Methods
Four hundred fifty patients underwent a living-donor liver transplant from January 2010 to June 2013. Fifty-four grafts with a GRWR less than 0.8 were used and made up the study group. To eliminate huge difference in sample size, 132 patients were selected randomly as the control group from the remaining 396 patients with adequate GRWR. (Every third patient was selected to be part of the control group).
For donor selection in Taiwan, we accept donors between 18 and 55 years of age, up to third-degree relatives. We accepted graft > 35% of standard liver volume. Whole liver volumes were calculated from the equation: liver volume (mL) = 706.2 × body surface area (m2) + 2.4. All donors in this study has steatosis < 10% on computed tomographic estimation, according to right or left graft recovered according to standard living donor technique.
In right lobe recovery, we generally keep the middle hepatic vein with the remnant liver of the donor. Segment V and/or segment VIII veins were reconstructed if there was a significant amount of congestion or backflow at the back table. Otherwise they were ligated. In left lobe grafts, middle hepatic vein was taken with the graft.
We used splenic artery ligation or splenectomy as portal flow modulation, if the portal flow after reperfusion was greater than 250 mL/min/100 g associated with the hepatic arterial flow less than 100 mL/min in low GRWR grafts (considering the hepatic artery buffer response as a causative factor for graft dysfunction). We did not do portal flow modulation in adequate GRWR grafts, so we did not do portal flow modulation in the control group. Generally, we prefer splenic artery ligation. However, splenectomy was used when the patient had pancytopenia with a cause of hepatitis C virus (HCV), particularly if we anticipate the need of postoperative interferon therapy.
Portal flow was measured by intraoperative Doppler analysis immediately after reperfusion (portal flow day 0). We do ultrasonographic Doppler examinations immediately after reperfusion, after abdominal closure, and daily for first 5 postoperative days. Standard hepatic vascular protocol (Acuson 128 scanner [Acuson, Mountain View, CA, USA] with 7.0 or 4.0 MHz scanners) was followed. The flow velocity and cross-sectional area of the portal volume were measured with Doppler mode. Portal flow was calculated by flow velocity and portal vein diameter just beyond the portal vein anastomosis.
Small-for-size syndrome was defined according to Clavien and Kyushu University definitions. Parameters like Model for End-Stage Liver Disease (MELD), cause, portal vein flow, age, and sex were compared as causative factors for small-for-size syndrome between the 2 groups. Statistical analyses were performed with SPSS software (SPSS: An IBM Company, version 21.0, IBM Corporation, Armonk, NY, USA). Categorical data were compared using chi-square test or Fisher t test, and continuous data were evaluated using Mann-Whitney U test. Multivariate analyses were done using logistic regression method. Receiving operating characteristic curves were prepared with 95% confidence interval. All grafts had < 10% steatosis by computed tomographic estimation.
Early graft dysfunction was defined according to definitions given for small-for-size syndrome by Dahm and associates and Sanefuji university group.4,5 Dahm and associates defined small-for-size graft dysfunction as dysfunction of small partial liver graft (GRWR < 0.8%) during the first postoperative week after the exclusion of other causes. They have defined graft dysfunction as the presence of 2 of the following on 3 consecutive days: bilirubin >100 μmol/L, international normalized ratio > 2, encephalopathy grade 3 or 4. Their exclusion criteria were technical (eg, arterial or portal occlusion, outflow congestion, bile leak), immunologic (eg, rejection), or infectious (eg, cholangitis, sepsis). Sanefuji and associates5 defined small-for-size syndrome as total bilirubin concentration > 86.2 μmol/L (5 mg/dL) on the postoperative day 14 in the absence of any other definitive causes of cholestasis, such as technical problems or immunological and infectious conditions with presence of intractable ascites (defined as daily production of ascites of > 1000 mL on postoperative day 14 or > 500 mL on postoperative day 28). We preferred to use the term graft dysfunction instead of small-for-size syndrome in the normal GRWR group.
We tried to analyze factors responsible for small-for-size syndrome (graft dysfunction in the normal size group) in both low GRWR group and normal GRWR group. (Total 22 patients developed early graft dysfunction in both groups.) Categorical variables were analyzed using the chi-square test and continuous variables with Mann-Whitney U test. P values < .05 were considered as statistically significant.
On univariate analyses, HCV P = .047) and hepatitis B virus (HBV) and hepatitis C carcinoma (HCC) (P = .047 each) cause, MELD score greater than 13 (P = .028), portal flow as a continuous variable (P = .001) was significantly associated with early graft dysfunction (Tables 1 and 2). There were no significant differences in complications between Low GRWR group and control group (Table 3).
Multivariate analysis was done using multivariate logistic regression method. Only portal vein flow as a continuous variable was significantly associated with graft dysfunction (P = .002; an odds ratio 1.012; a 95% confidence interval: 1.004-1.021). On receiver operating characteristic curve area under the receiver operating characteristic curve was 0.719, with the P value of .001 with a 95% confidence interval: 0.627-0.811 (Figure 1), portal flow greater than 190 was showed the highest sensitivity and specificity on receiver operating characteristic analysis with area under the receiver operating characteristic curve of 0.741 with a 95% confidence interval: 0.532-0.799 (Figure 2). Note that GRWR was not a significant predictor for graft dysfunction fulfilling the small-for-size syndrome definition at any stage.
It is established that portal vein hemodynamics is the main factor responsible for small-for-size syndrome. Marcos and associates,6 showed “extreme imbalance in donor-recipient portal vein flow (PVF) has been recorded successively in, characterized by increased portal venous inflow and a relatively reduced supply from the hepatic artery.” Lautt and associates7 suggest adenosine washout hypothesis was responsible for this phenomenon. A constant release of adenosine, a vasodilator substance, among the hepatic arterioles and portal venules maintains balanced inflows. Increased portal flow decreases local adenosine concentrations resulting in hepatic artery branch constriction and a reduction in arterial flow, which is responsible for graft dysfunction in small-for-size syndrome. Troisi and associates8 showed that portal flow inflow modulation when portal flow is greater than 250 mL/min/100 g of the graft significantly improves patient and graft survival and reverse the histologic changes of hepatic injury in small-for-size grafts, based on various studies mentioned above, we adopted portal flow modulation when portal flow immediately after reperfusion is greater than 250 mL/min/100 g. Asencio and associates9 formulated a hypothesis that development of the small-for-size syndrome is not exclusively determined by the ratio to the mass of the liver remnant (or graft) to body weight, but instead, is strictly determined by the hemodynamic parameters of the hepatic circulation. Kaido and associates 3 suggested that with a low GRWR graft up to 0.6 with portal pressure control could be used with acceptable recipient survival rates and donor complication rates.3
We used portal flow instead of portal pressure in our study because of the findings of Troisi and associates2,8 and to test hypothesis developed by Asencio9 and also because Chan and associates10 showed that the graft portal inflow, when expressed in volume per 100 g of graft weight per minute, had a linear correlation with the recipient portal pressure before hepatectomy of the native liver (and was weaker in the anhepatic phase) but not after graft implantation. They showed that portal flow postreperfusion was not correlated with portal pressure.
Troisi and associates2 suggested that following reperfusion of a partial graft, hemodynamic changes were more pronounced than those occurring in deceased-donor liver transplants, with an inverse correlation to the GRWR. Overall mean PVF values are increased at least twice the donor values (2-7 times more according to our data), and the portal contribution may raise the total graft inflow. They also suggested that “hyperperfusion of small grafts is thought to be one of the main causes of posttransplant graft dysfunction (small-for-size syndrome), and poor graft outcome has been reported in patients with definitive values of more than 260 mL/min/100 g. Chan and associates11 also suggested portal hyperperfusion as an indication of possible portal hypertension. Portal manometry can be done with cannulation of the inferior mesenteric vein or by direct puncture of the portal vein. The latter, however, does not allow for continuous monitoring, and portal pressure can vary with central venous pressure, which in turn can vary with the positions of the patient and the catheter tip.11
Sainz-Barriga and associates12 found no correlation between postreperfusion portal pressure and portal flow. They suggested the need for portal flow modulation does not require elevation of both. They also described that portal flow modulation technique has the direct effect on portal flow reduction. In their study, 25% of patients had portal vein pressure greater than 20 mg and portal flow of 90 mL/min/100 mg, so that if portal flow modulation is practiced on the basis of portal pressure, there is a high risk of portal flow hypoperfusion in these patients. Based on these findings, we took portal flow measured by Doppler as a guide for portal flow modulation.
However, no study to our knowledge has shown the effect portal flow modification in development for small-for-size syndrome in the early postoperative period strictly according to definitions. Recently, we have accepted grafts if the standard liver volume is greater than 35% with GRWR ≥ 0.6 with portal vein modulation if portal flow is greater than 250 mL/min/100 g. We perform either splenic artery ligation or splenectomy as described earlier. As our results showed that splenic artery ligation was sufficient to decrease portal flow in our study group, 2 patients with HCV etiology because of associated pancytopenia undergone splenectomy, portal flow in these 2 patients also reduced adequately. Troisi and associates2 also mentioned that splenic artery ligation is adequate for portal flow inflow modulation and also helps in maintaining adequate hepatic artery inflow as well as prevents thrombocytopenia. They mentioned that portocaval shunt is only required when portal flow is > 500 mL/min/100 g. We also achieved adequate reduction in the portal flow by splenic artery ligation or splenectomy in case of HCV patients with pancytopenia.
Out of 54 patients with low GRWR grafts, only 6 patients developed
small-for-size syndrome. Mean GRWR in low GRWR group was 0.71. There was no
significant difference in early graft dysfunction in low GRWR and adequate GRWR
group (P = .762). We use the term early graft dysfunction in adequate
GRWR group when they fulfill the definitions of small-for-size syndrome, but as
they were in normal GRWR group small-for-size syndrome is not the suitable term.
Graft volume/standard liver volume also did not achieve statistical significance
in predicting graft dysfunction (P = .530). As shown earlier, only 2 of
10 patients underwent portal flow modulation when portal flow was greater than
250 mL/min/100 g developed small-for-size syndrome. Both the patients in normal
for size group who had portal flow greater than 250 mL/min/100 g without portal
flow modulation showed features of graft dysfunction, which was statistically
(P = .013). There was no statistically significant difference in complications between the 2 groups as shown in Table 3.
We evaluated cutoff value of 190 mL/min because the 2 grafts developed small-for-size syndrome even after portal flow modulation had portal flow greater than 190 mL/min, and mean portal flow also was greater than 190 mL in graft dysfunction group and also on receiver operating characteristic analysis portal flow greater than 190 mL/min was showing highest sensitivity and specificity. After multivariate logistic regression on portal flow independently predicted graft dysfunction in both the group with the P value of .002, this suggests that PVF, and not the GRWR or GV/SLV, is responsible for small-for-size syndrome. Minimum follow-up was 30 days in our study, as we wanted to study early graft dysfunction fulfilling small of size syndrome ranging from 30 days to 2 years. No graft loss, retransplant and mortality were noted in low GRWR group. Graft/recipient's body weight ratio and GV/SLV can still be useful in preoperative evaluation in selecting the donor, but the intraoperative decision should be based on portal flow only. By using portal flow inflow modulation technique smaller grafts and left lobe can be used more with adequate result and increasing donor safety. In the univariate analysis, MELD was also associated significantly with early graft dysfunction. In our study portal flow greater than 190 mL/min/100 g. Statistically predicted development of early graft dysfunction regardless of GRWR status. This shows lowering cutoff for portal flow modulation will help prevent early graft dysfunction. However, further studies are warranted to determine the absolute cutoff value of portal flow to prevent small-for-size syndrome as well as maintain adequate portal perfusion.
This study has several limitations, as it was retrospective study inherent limitation was retrospective studies were there. As we did portal flow modification in all patients with portal flow greater than 250 mL/min/100 g. We did not have enough subgroup to compare an outcome with or without portal flow modification when portal flow was greater than 250 mL/min/100 g. Further studies are required to evaluate interrelationship of portal flow and portal pressure.
In conclusion, using smaller graft and left lobe with portal flow modification is a reasonably strategy to increase donor pool and donor safety with decreasing incident of small-for-size syndrome. Portal flow and not GRWR or GV/SLV predicts small-for-size syndrome or early graft dysfunction. Cutoff limits of portal flow should be lowered to decrease the incidence of small-for-size syndrome. However, further studies are warranted to determine the absolute cutoff value.
Volume : 12
Issue : 5
Pages : 437 - 442
DOI : 10.6002/ect.2014.0087
From the 1Department of Hepatobiliary Surgery and
Liver Transplantation, Continental Hospital, Hyderabad, India; and the 2Liver
Transplant Program, Kaohsiung Chang Gung Memorial Hospital, Kaohsiung, Taiwan
Acknowledgements: The authors have no conflicts of interest to disclose, and there was no funding for this study.
Corresponding author: Dr. Bhavin B. Vasavada, Junior Consultant, Hepatobiliary and Liver Transplantation, Continental Hospital, Hyderabad, India
Phone: +91 767 493 4580
Fax: +91 040 2670 0000
Table 1. Patient Characteristics
Table 2. Causative Factors
Table 3. Complications in Low GRWR Group Versus Control Group
Figure 1. Receiver Operating Curve of Portal Flow as a Continuous Factor
Figure 2. Receiver Operating Curve of Portal Flow Greater Than 190 mL/min/100 g