Objectives: We sought to identify the risk factors associated with the early mortality after a living-donor liver transplant.
Materials and Methods: Two hundred eighteen patients were recruited in this study. Potential risk factors were analyzed using univariate and multivariate analyses. A C statistic equivalent to the area under the receiver operating characteristic curve was used to assess the ability of the model to predict mortality risk during the first 3 months after a living-donor liver transplant.
Results: Twenty-six recipients died within the first 3 months after a living-donor liver transplant. On a multivariate analysis, intraoperative allogeneic red blood cell transfusion and the preoperative creatinine levels were independently associated with early postoperative mortality. A prognostic model was proposed in this study (early mortality risk score = 0.107 × intraoperative allogeneic red blood cells transfusion [U] + 0.005 × preoperative creatinine concentration [μmol/L]). Three-month survival rates of patients with high and low scores were 69.8% and 95.5% (P < .001).
Conclusions: Transfusion of intraoperative allogeneic red blood cell and preoperative creatinine levels are associated with the early mortality after living-donor liver transplant. A model to predict early mortality after a living-donor liver transplant based on these risk factors was proposed in this study.
Key words : Model for end-stage liver disease, Serum sodium, Predictors
Liver transplant is widely accepted as the curative treatment for patients with end-stage liver disease. However, worldwide scarcity of deceased-donor liver grafts limits its management. Living-donor liver transplant is a useful alternative to deceased-donor liver transplant, and has saved many lives in past 20 years. As a complex surgical procedure, many factors may negatively influence the outcomes of patients undergoing living-donor liver transplant, especially during the early postoperative period.1 Previous investigations suggested the early mortality after liver transplant was about 5% to 10%.2 However, which factors contribute to early death after liver transplant has not been well established. Although previous investigations suggest the severity of liver disease assessed by Model for End-Stage Liver Disease score is useful to predict early mortality after liver transplant, it remains controversial.3-6 Here, we attempt to identify risk factors associated with early mortality after a living-donor liver transplant.
Materials and Methods
All adult patients who underwent a living-donor liver transplant at our center between 2005 and 2012 were recruited into the present study. Patients receiving a dual graft living-donor liver transplant, repeat transplant, or a combined liver and kidney transplant were excluded. All transplants and the current study were approved by the ethics committee of West China Hospital, Sichuan University. All of the protocols conformed with the ethical guidelines of the 1975 Helsinki Declaration. Written informed consent was obtained from all subjects.
Liver donors were healthy relatives within the third degree of consanguinity with a compatible ABO blood type. Serologic testing for viral hepatitis and human immunodeficiency virus antibodies and testing for other acute or chronic diseases had to prove negative. We used a volumetric computed tomography with contrast to evaluate the donor’s liver. The right hepatic lobe without the middle hepatic vein of the donor’s liver had to be at least 0.8% of the weight of the recipient’s, and the remnant liver had to be at least 40% for the donors.
A detailed surgical procedure has been described previously.7 Briefly, after induction of general anaesthesia, surgery for donors was done through a right subcostal incision with an extension to the upper midline. After abdominal exploration, a liver biopsy was performed to assess the amount of steatosis present. After the cholecystectomy, an intraoperative cholangiogram under fluoroscopy via the cystic duct was done to evaluate the biliary anatomy. Liver transection was performed using a Cavitron ultrasonic surgical aspirator (Valleylab [a division of Tyco Healthcare Group LP], Boulder CO, USA) without inflow occlusion. At the end of the liver parenchymal transection, the right bile duct, hepatic artery, and portal vein were divided. All hepatic veins larger than 0.5 cm were preserved until recovery for potential anastomosis in the recipient. The falciform ligament was reconstructed, and the stumps of the hepatic vein and portal vein were closed with continuous nonabsorbable sutures. Before closure, a drain was inserted into the right subphrenic cavity.
All grafts were preserved and flushed with University of Wisconsin solution. Reconstruction of the hepatic artery was done after reperfusion. A jump graft to the aorta was done using the recipient’s saphenous vein if recipient’s hepatic artery was inadequate. A duct-to-duct anastomosis was the first choice for bile duct reconstruction. A bile duct stent was not used in any patients. If the recipient’s bile duct was inadequate, a Roux-en-Y hepaticojejunostomy was performed. A veno-venous bypass was not used in any recipients.
Postoperative immunosuppression consisted of a calcineurin inhibitor (tacrolimus or cyclosporine), mycophenolate mofetil, and a steroid. Tacrolimus was usually initiated within the first 24 hours after surgery. For patients with severe renal dysfunction, the calcineurin inhibitor was replaced with sirolimus. Calcineurin inhibitors were administrated to these recipients when their renal function had stabilized. Steroids were withdrawn as soon as possible. Rejection was treated with steroid pulse therapy. Lamivudine and hepatitis B immune globulin were used after the liver transplant to prevent recurrence of hepatitis B virus in patients with positive pretransplant HBsAg. Hepatitis B immune globulin was used during transplant for hepatitis B virus patients.
The recipient’s age, gender, indication for liver transplant, Model for End-Stage Liver Disease score, total bilirubin, international normalized ratio, creatinine, albumin, serum sodium, body mass index, preoperative hemoglobin, white blood cell, and platelet levels were collected. The donor’s age, gender, and body mass index also were collected. Intraoperative variables including graft-to-recipient body weight ratio, blood loss, and allogeneic red blood cell transfusion were retrospectively reviewed. Model for End-Stage Liver Disease and Model for End-Stage Liver Disease-based models were calculated with the following formulas: Model for End-Stage Liver Disease score = 9.57 × Ln creatinine (mg/dL) + 11.2 × Ln international normalized ratio + 3.78 × Ln bilirubin (mg/dL) + 6.438; blood transfusions were based on the clinical evaluation and laboratory findings. Target hemoglobin level was not less than 7.0 g/dL.9 Fresh frozen plasma was used when the international normalized ratio was greater than 1.5.9 Early mortality was defined as recipients dying during the first 3 months after a liver transplant.10 Estimated glomerular filtration rate was calculated using the Cockcroft-Gault method: estimated glomerular filtration rate = (140-age) × weight/creatinine × 72 × 0.85 (if female).11
Survival of the recipient was estimated using the life-table method and compared using the Kaplan-Meier method with the log-rank test. Continuous variables were presented as means ± SD and compared using a 1-way analysis of variance, whereas categorical variables were analyzed using the chi-square test or Fisher exact test. The Cox regression analysis was used to identify the variables that were independently associated with early postoperative mortality. Significant variables, those with a P value less than 0.1 in the univariate analyses, were subsequently included in the multivariate analysis. To identify the individual risk of early postoperative mortality in a given patient, an equation was developed that identified independent risk factors: F(x) =β1x1 + β2x2 +…+ βixi, where xi is the identified independent risk factor, and βi is the regression coefficient. Sensitivity and specificity were estimated using the receiver operating characteristic curve. Statistical analyses were performed with SPSS software (SPSS: An IBM Company, version 16.0, IBM Corporation, Armonk, NY, USA). P values less than .05 were considered statistically significant.
Two hundred eighteen recipients were included in the current study. As shown in Table 1, the indications for liver transplant included hepatitis B virus cirrhosis (n = 131), hepatocellular carcinoma (n =69), alcoholic cirrhosis (n = 4), hepatitis C virus cirrhosis (n = 4), hepatic hydatidosis (n = 3), hepatolithiasis (n = 2), trauma (n = 1), huge hepatic hemangioma (n = 1), polycystic liver (n = 1), autoimmune hepatitis (n = 1), and Budd-Chiari syndrome (n = 1). The mean age of the recipients was 42.75 ± 8.79 years, whereas the mean age of the donors was 36.39 ± 9.85 years. The original Model for End-Stage Liver Disease scores ranged from less than 6 to more than 40, with a mean of 15.80 ± 9.73. The mean graft-to-recipient body weight ratio was 0.95% ± 0.18%. During the follow-up, 26 recipients (11.93%) died. The causes of death were multiple organ dysfunction syndrome (n = 14), infection (n = 7), renal failure (n = 3), and bleeding (n = 2). The overall 3-month survival rate was 88.1%.
Risk factors possibly associated with 3-month mortality after a living-donor liver transplant were investigated in the univariate analysis. As shown in Table 2, significant factors predicting mortality in the first 3 months after a living-donor liver transplant were preoperative total bilirubin level, preoperative hyponatremia, preoperative creatinine level, intraoperative blood loss, intraoperative red blood cells transfusion, and preoperative estimated glomerular filtration rate. These potential risk factors were entered into a multivariate analysis to determine their independent significance. On multivariate analysis, only preoperative creatinine and intraoperative red blood cell transfusion were identified as independent predictors of a poor prognosis (Table 3).
After multivariate analysis, the regression coefficients listed in Table 3, and the risk factors independently associated with the 3-month mortality after a living-donor liver transplant, were used to generate the following formula: Early mortality risk = 0.107 × intraoperative allogeneic red blood cells transfusion (U) + 0.005 × preoperative creatinine concentration (μmol/L). The receiver operating characteristic curve determined that the best cutoffs for early mortality were values greater than 1.3, with a sensitivity of 76.9%, and a specificity of 75.5% (Figure 1). As shown in Figure 2, the 3-month survival rates for patients in the current study with high (risk score > 1.3) and low (risk score ≤ 1.3) early mortality scores were 69.8% and 95.5%. A statistically significant difference was observed (P < .001).
Living-donor liver transplant is a complex surgical procedure. Many factors negatively influence the outcomes of patients undergoing a living-donor liver transplant, especially during the early postoperative period. Our study confirmed that the preoperative creatinine concentrations and intraoperative red blood cell transfusions were independent risk factors for the first 3-month mortality after a living-donor liver transplant.
This study shows that preoperative creatinine concentrations are independent risk factors for mortality. Model for End-Stage Liver Disease score includes the patient’s preoperative creatinine level and suggests that patients on waiting lists with a high creatinine levels have a higher risk of mortality. Our results further confirmed that the risk of early postoperative mortality is increased with an increasing preoperative creatinine concentration. Renal dysfunction is common and present in 10% to 20% of patients undergoing a liver transplant.12 Previous investigations have suggested that renal dysfunction is associated with a prolonged hospital and intensive care unit stay, more postoperative infections, and poor survival rates.13,14 Dellon and associates15 have suggested that combined liver and kidney transplant may be the best treatment for older patients who need preoperative renal replacement therapy. The liver transplant procedure is difficult and is associated with major hemodynamic changes. Many variables, such as massive blood loss and secondary hypotension, a prolonged anhepatic phase, and intraoperative electrolyte and acid-base disorders, may cause acute kidney injury or further impairment of the renal function.16-18
Several investigations have reported that massive intraoperative allogeneic red blood cell transfusions have a negative effect on liver transplant recipient outcomes.19-21 Ramos and associates22 confirmed that massive red blood cell transfusions are associated with longer hospital stays, and decreased postoperative survival for liver transplant recipients. Benson and associates23 suggest that red blood cells transfusion contributes to postoperative infections in a dose-dependent manner. Hendriks and associates24 report that the requirement for intraoperative blood products is a risk factor for postoperative reintervention after a liver transplant. Intraoperative red blood cell transfusion is an independent risk factor for early postoperative mortality after a liver transplant in the present study. This result suggests that postoperative outcomes of liver transplant depend partly on the intraoperative management.
Model for End-Stage Liver Disease scores showed poor predictive power for early postoperative mortality after a living-donor liver transplant in the present study. However, some investigations suggest that a high preoperative Model for End-Stage Liver Disease score is associated with poor postoperative outcomes.25,26 Previous studies confirm the biochemical variables of recipients who recovered rapidly after living-donor liver transplant and have shown these returned to similar levels during the first postoperative month among groups with different Model for End-Stage Liver Disease scores.27 Jacob and associates28 suggested that some unpredictable variables during the transplant, in addition to the severity of disease at the time of transplant, may negative influence the postoperative outcomes after liver transplant. Our results suggest that intraoperative red blood cells transfusion is associated with the early postoperative mortality. However, Model for End-Stage Liver Disease score and Model for End-Stage Liver Disease-based models all are based on the preoperative variables and include no intraoperative variables. Accordingly, Model for End-Stage Liver Disease scores and Model for End-Stage Liver Disease-based models did not show predictive power in the present study.
There are limitations in the present study. We acknowledge that including intraoperative variables (red blood cells transfusion) in our model may limit its use as a preoperative model for decision making. However, we suggest that intraoperative variables must be taken into account when analyzing the risk factors that negatively influence postoperative outcomes. Many investigators have confirmed the relation of intraoperative events and postoperative outcomes.29,30 Our model is useful for patients with preoperative renal dysfunction. Although some reports have indicated that preoperative renal dysfunction is associated with poor postoperative outcomes, it seems that the reduction of intraoperative red blood cells transfusion may minimize the mortality risk for these sicker patients. Moreover, previous investigators also have suggested that red blood cell transfusions may cause acute kidney injury and aggravate renal dysfunction.31
Intraoperative allogeneic red blood cells transfusion and the preoperative creatinine level are independent risk factors for early postoperative mortality. A model based on intraoperative allogeneic red blood cells transfusion and preoperative creatinine level, which was proposed in the present study, shows good prognostic power for predicting early postoperative mortality after a living-donor liver transplant.
Volume : 13
Issue : 1
Pages : 62 - 67
DOI : 10.6002/ect.2014.0031
From the Division of Liver Transplantation, West China Hospital of Sichuan
University, Chengdu (610041), China
Acknowledgements: Chuan Li and Wen Tian-Fu Wen proposed the study. Chuan Li collected and analyzed the data. All authors contributed to the surgery and the relevant discussion. We would like to express our appreciation to the China Liver Transplantation Registry for performing the data collection. This work was supported by a grant from the National Science and Major Technology Project of China (2012ZX10002-016 and 2012ZX10002-017). The authors have no conflicts of interest to declare.
Corresponding author: Tian-Fu Wen, Division of Liver Transplantation, West China Hospital of Sichuan University, Chengdu (610041), China
Phone: +86 189 8060 1471
Table 1. Demographic Data of Donors and Recipients
Table 2. Univariate Analysis of the Variables Used To Predict 3-Month Mortality After Living-Donor Liver Transplant
Table 3. Demographic Data of Donors and Recipients
Figure 1. Receiver Operating Characteristic Curve for Early Mortality Risk
Figure 2. Survival Curves of Patients With Low and High Early Mortality Risk Scores (P < .001)