Objectives: The outcome of children who had living-donor liver transplant was analyzed according to their status before transplant, and we analyzed the outcome of critically ill patients.
Materials and Methods: This was a retrospective analysis of children who received primary living-donor liver transplant at Kyoto University Hospital. According to the criteria of the United Network for Organ Sharing, we divided patients into 3 groups: Group A patients had been admitted to the intensive care unit before living-donor liver transplant; Group B patients were hospitalized but did not require intensive care unit stay; and Group C patients were living at home and underwent elective transplant.
Results: A total 685 patients met inclusion criteria. Children in Group A were younger than Group B and received liver grafts from younger donors than Group B and C. Group A patients had marked impairment in liver and renal function and coagulation profile and needed higher volumes of fresh frozen plasma transfusions. Group A patients had significantly worse outcomes and early patient death than the other group; Group A patient survival was 68.3%, 63.2%, 60.1%, and 56.1% at 1, 5, 10, and 15 years after living-donor liver transplant (P < .0001). Group A had worse graft survival than other groups (P < .0001), and Group A graft survival was 68.3%, 65.9%, 54.1%, and 49.9% at 1, 5, 10, and 15 years. Low gamma-glutamyl transpeptidase was an independent risk factor for patient death in Group A (hazard ratio, 1.004; 95% confidence interval, 1.0-1.007) (P < .05). Group A patients had a higher rate of multidrug-resistant hospital-acquired infections.
Conclusions: Children who were admitted to the intensive care unit prior to living-donor liver transplant had marked impairment of pretransplant laboratory parameters and worse outcome than other groups.
Key words : End-stage liver disease, Infection, Intensive care unit, Pediatric, Risk factors
Although better immunosuppressive agents and surgical techniques partially account for increasing survival with living-donor liver transplant (LDLT), advances in pre- and postoperative care also have helped improve survival. Liver transplant has been accepted as the standard therapy for patients with end-stage liver disease. As a result, there has been a substantial increase in the number of patients placed on waiting lists for transplant.1,2 Infection after pediatric LDLT is a major cause of morbidity and mortality, especially in critically-ill patients, as described in several studies.3-6
To our knowledge, the outcome of severely ill children after LDLT has not been adequately reported. Therefore, the objective of this retrospective cohort study was to compare the outcomes of critically ill children who underwent LDLT with the outcomes of patients who were less ill, as defined by the United Network for Organ Sharing criteria.7 Patients were considered severely ill when they had been admitted to an intensive care unit (ICU) before the liver transplant (termed ICU-bound); patients who had been admitted to a hospital but not an ICU, and patients who had been living at home at the time of transplant, were considered less ill. Patient survival, graft survival, and risk factors for survival and infectious complications were assessed.
Materials and Methods
This study included children who received primary LDLT at Kyoto University Hospital between June 1990 and April 2011. Inclusion criteria were primary liver transplant with standard techniques for end-stage liver disease, both chronic and acute fulminant hepatic failure. Patients who had retransplant and auxiliary liver transplants were excluded. According to United Network for Organ Sharing criteria, we divided patients into 3 groups: Group A patients had been admitted to the ICU before LDLT (ICU-bound); Group B patients were hospitalized but did not require ICU stay; and Group C patients were living at home and underwent elective transplant. All clinical and laboratory data were collected from the patient charts, and patient and graft survival were assessed.
Techniques for donor and recipient operations have been described previously.8,9 The left lateral segment was the primary choice. However, when the estimated graft recipient weight ratio (GRWR) was > 4%, a monosegment graft was used.10,11 For larger recipients, graft selection was extended to the left lobe11 and right lobe12 according to GRWR and the residual liver volume in the donor after hepatectomy.
Immunosuppression consisted of combination therapy with tacrolimus and steroids.13 The tacrolimus started orally from the day before the operation, then continued postoperatively. Target tacrolimus trough serum levels initially were > 10 ng/mL, decreasing gradually to 6 to 8 ng/mL a few months after LDLT. Methylprednisolone therapy was used for induction and switched to oral prednisolone therapy 1 week after LDLT. Steroid therapy was routinely tapered by 3 to 6 months after transplant when graft function was maintained. In cases of ABO-incompatible LDLT, additional immunosuppressants and preconditioning regimens were given to inhibit humoral rejection; treatment included prostaglandin E1, cyclophosphamide, azathioprine, mycophenolate mofetil, and plasma exchange.14
All clinical and laboratory data were retrieved from the patient charts. The values used for analysis were from the last records before LDLT.
Risk factors for patient survival
We evaluated the risk potential of several preoperative, operative, and postoperative variables. The preoperative variables included age at LDLT, sex, preoperative clinical status, presence of preoperative infection, ascites, laboratory variables (white blood cell count, C-reactive protein, electrolytes [sodium, potassium, calcium, phosphate, magnesium], liver function tests [aspartate aminotransferase, alanine aminotransferase, total bilirubin, gamma-glutamyl transpeptidase (GGTP), and albumin], coagulation factors [international normalized ratio, antithrombin ІІІ level, platelet count, and prothrombin time], renal function tests [blood urea nitrogen and creatinine levels]), preoperative hospital stay, ABO-mismatching, and pediatric end-stage liver disease score. Operative variables included operative time, cold ischemia time, warm ischemia time, blood or blood product transfusion (packed red blood cells), fresh frozen plasma, platelets, fluids, 5% albumin), blood loss, and GRWR. Postoperative variables included surgical complications (intraabdominal hemorrhage, bile leakage, or intestinal perforation), repeat surgery, (duration of insertion of intravascular catheter, intraabdominal drainage, or bile drainage), or graft dysfunction. Graft dysfunction was defined as persistent abnormal liver function with serum aminotransferase levels at 2 to 3 times normal, with or without elevated bilirubin level, and abnormal biopsy findings. Data about concomitant rejection or administration of steroid pulse at the time of infection also were evaluated.
The patients received flomoxef, an oxacephem antibiotic, 1 hour before the operation, and this was continued for 72 hours after surgery. Trimethoprim and sulfamethoxazole were administered once daily as prophylaxis against Pneumocystis. Miconazole was administered for 7 days after transplant as antifungal prophylaxis. Antiviral prophylaxis, including ganciclovir, was not administered except in seronegative recipients who received an allograft from cytomegalovirus-seropositive donors.
The protocol for the evaluation of infections during the posttransplant hospital stay included testing for endotoxin (gram-negative bacteria), D-glucan (Pneumocystis carinii and fungi except Cryptococcus neoformans), cytomegalovirus DNA (polymerase chain reaction), and Epstein-Barr virus DNA (semiquantitative polymerase chain reaction) in peripheral blood once weekly and additionally when there was a suspicion of infection, and culture of blood, any wound discharge, urine, stool, pharyngeal swabs, sputum, and bile ≥ twice weekly.
Definition of infections
Bacterial and fungal infections were defined using the criteria proposed by the Centers for Disease Control and Prevention.15,16 Primary bacteremia was defined as bacteremia with no physical, radiographic, or pathologic evidence of a definite infection source. Catheter-related blood stream infection was defined when the same organism was cultured from the catheter tip and blood culture.17 Secondary bacteremia was defined when blood cultures and cultures of samples collected from the suggested infection site showed the same organism. Surgical site infection included cholangitis, peritonitis, intraabdominal abscess, and wound infection.18
Confirmed fungal infections were diagnosed according to positive culture and body temperature > 38°C. Probable or possible fungal infections were diagnosed when there was a positive finding for D-glucan antigen or Aspergillus antigen, or when the patient was febrile despite the administration of broad-spectrum antibiotics.19
Viral infections were diagnosed by clinical findings and the detection of viral DNA or RNA fragments by quantitative real-time polymerase chain reaction.20 In addition, systemic lymphadenopathy was assessed, and biopsy of the lymph nodes at the body surface, including the axillary and inguinal lymph nodes, was performed to exclude posttransplant lymphoproliferative disorder.
Postoperative treatment of infection
Bacterial infection initially was treated empirically with vancomycin (40-50 mg/kg/d) (recommended trough level < 10 mg/L), and antibiotics were adjusted or changed according to the identified organism. In patients who had sepsis, the tacrolimus dose was decreased or completely withdrawn, and liver biopsy was performed to exclude the possibility of rejection.
Fungal infection was treated with antifungal agents such as miconazole or fluconazole. Amphotericin B was added when there was no improvement. Wide spectrum antibiotics were stopped. The immunosuppression drugs were decreased and liver biopsy was performed. In addition, micafungin sodium (50-150 mg/d) was administered.
Treatment of viral infection varied. Cytomegalovirus disease was treated with ganciclovir (10 mg/kg/d intravenous for 2 weeks, and tapered to 5 mg/kg/d) until clinical resolution occurred, and tacrolimus dosage was minimized. Epstein-Barr virus disease was treated with acyclovir (60 mg/kg/d) or vidarabine (10 mg/kg/d) and reduction in the dosage or complete cessation of tacrolimus. When Epstein-Barr virus DNA accounted for > 105 copies/μg DNA, tacrolimus was temporary stopped until Epstein-Barr virus DNA level decreased to < 104 copies/μg DNA.
Kruskal-Wallis test was used to compare continuous variables between the 3 groups, followed by pairwise comparisons when a significant difference was identified. Chi-square test was used to evaluate associations between the 3 groups for categoric variables. Overall patient survival was described by Kaplan-Meier method and compared using log-rank test. The outcome was defined as graft failure or patient death after LDLT, and analysis for risk factors was done using Cox proportional hazards regression model. Hazard ratios and 95% confidence intervals were assessed. Data analysis was performed with statistical software (SPSS for Windows, Version 16.0, SPSS Inc., Chicago, IL, USA). Values of P ≤ .05 were considered significant.
In 1354 subjects who underwent LDLT in the study sample (from June 1990 to April 2011), a total 685 patients met the inclusion criteria (mean follow-up, 7.52 ± 5.13 y). The mean age for all patients was 4.06 ± 4.61 years (median, 1.66 y; range, 0.12-17.87 y), and 414 recipients were females (60.4%). Group A included 74 patients who had been admitted to the ICU before LDLT; Group B had 354 children who were hospitalized but did not require ICU admission; and Group C included 257 patients who were living at home and underwent an elective transplant. There were significant differences between patients of the 3 groups (Table 1). In Group A, children were younger than Group B and received liver grafts from younger donors than Group B and C. Biliary atresia was the main indication for LDLT in all groups, but the proportion of biliary atresia was lower in Group A (31%) than Group B or C (Table 1). Fulminant hepatic failure constituted 32% Group A patients. Group A patients had marked impairment in liver and renal function and coagulation status as indicated by the high bilirubin, creatinine, blood urea nitrogen, prothrombin time, and international normalized ratio. However, there were comparable values for albumin and liver enzyme levels between the groups (Table 1). Intraoperative variables and postoperative complications were studied, and we observed that Group A had significantly higher mean transfusion volume of fresh frozen plasma than Group B or C (Table 2).
Patient and graft survival
Patients who received ICU care before LDLT (Group A) had significantly worse outcome than the other groups. Group A had patient survival 68.3%, 63.2%, 60.1%, and 56.1% at 1, 5, 10, and 15 years after LDLT (P = .001) (Figure 1). Group B and C had better patient survival than Group A, and patient survival was similar for Group B and C (P = .799). In Group B, patient survival was 87.6%, 85.9%, 82.7%, and 81.4% at 1, 5, 10, and 15 years; in Group C, patient survival was 89.4%, 87.1%, 82.4%, and 76.1% at 1, 5, 10, and 15 years (Figure 1). A remarkable feature of critically ill patients (Group A) was the poor outcome and the pattern of patient loss, because early patient loss accounted for decreased survival to 80% in the first month after LDLT.
Group A had markedly worse graft survival than other groups (P = .001) (Figure 2). Group A had graft survival 68.3%, 65.9%, 54.1%, and 49.9% at 1, 5, 10, and 15 years. In contrast, Group B had graft survival 89%, 85.9%, 80.1%, and 72.6% at 1, 5, 10, and 15 years, and Group C had graft survival 89.1%, 86.8%, 81.2%, and 73.8 % at 1, 5, 10, and 15 years (Figure 2).
Risk factors for patient survival after living-donor liver transplant
A univariate proportional hazards regression model was used to examine potential risk factors for association with patient survival. Univariate analysis revealed 4 potential significant risk factors for poor patient survival, including female sex, pretransplant platelet count, pretransplant lactate dehydrogenase level, and pretransplant GGTP level (P < .05) (Table 3).
The 4 potential risk factors derived from the univariate analysis were further assessed by multivariate analysis. The multivariate analysis revealed that low GGTP level was the only variable that had independent prognostic significance (Table 3).
Analysis of infection after pediatric living-donor liver transplant
The incidence and rate of hospital-acquired infection was 51% (2.5) in Group A, 61% (1.3) in Group B, and 54% (1.2) in Group C. Bacterial infection contributed to 64% infectious episodes in Group A, 63% in Group B, and 52% in Group C. Fungal infection was detected in 7.4% infectious episodes in Group A, 4% in Group B, and 6% in Group C. Viral infection was detected by polymerase chain reaction in 28% patients in Group A, 33% in Group B, and 42% in Group C (Tables 4, 5, and 6).
With respect to timing of infection, bacterial and fungal infections occurred at 13 ± 9 days in Group A, 12 ± 12 days in Group B, and 14 ± 12 days in Group C. Viral infections occurred at 18 ± 16 days in Group A, 22 ± 19 days in Group B, and 22 ± 13 days in Group C. The most common isolate was methicillin-resistant Staphylococcus aureus (MRSA) in Group A, Pseudomonas aeruginosa in Group B, and Enterococcus in Group C (Table 4).
This study is an important update about our experience with pediatric patients who underwent LDLT. It showed that GGTP had independent prognostic significance for recipient survival. Female sex, pretransplant platelet count, and lactate dehydrogenase were associated with poor patient survival. The ICU-bound children before transplant had the worst outcome after LDLT; they had higher infection rates and were complicated by fungal and multidrug resistant bacterial infections.
In this study, we examined the outcome of severely ill children who had been admitted to an ICU before liver transplant. Children who had been admitted to a hospital but not an ICU, and children who had been living at home before transplant, were considered less ill.
The ICU-bound children were younger than hospitalized children were and received liver grafts from younger donors than the other 2 groups. The most common indication for LDLT in most children was biliary atresia, but this was less frequent in ICU-bound children. Critically ill children had marked impairment in liver and renal function and coagulation profile; they received higher volumes of fresh frozen plasma transfusion and had a higher incidence of biliary leakage after LDLT. Patients who received ICU care before LDLT also had a significantly worse outcome than other groups.
The remarkable feature of critically ill patients was the poor outcome and also the pattern of patient loss, because early patient loss accounted for a decrease in survival to 80% in the first month after LDLT. In another study that examined patient and graft survival after LDLT in adult recipients who received ICU care before transplant, there was no difference in survival between the patient groups; in that study, factors that may have affected outcome such as age, sex, race, cause of disease, and cold ischemia time were similar between the patient groups.21 We observed that low pretransplant platelet count was associated with poor outcome in our children, and another study reported that low posttransplant platelet count may predict early posttransplant survival.22
The enzyme GGTP is a microsomal enzyme that is distributed widely in human tissues that are involved in secretory and absorptive processes, particularly the bile canaliculi. This enzyme also may help screen for biliary complications in patients who have orthotopic liver transplant.23 In this study, we observed that GGTP had independent prognostic significance for recipient survival. In another study after orthotopic liver transplant, an early increase in GGTP correlated with better outcomes.24 An experimental study found that GGTP is an early and sensitive marker in ethanol-induced liver injury in rats.25
We observed a high pretransplant bilirubin level in our cohort. Majority of our children who had biliary atresia were scheduled for LDLT due to failure of kasai operation. In consistent with previous 2 studies of children who had the Kasai operation for correction of biliary atresia, and those children in the pretransplant scenario exhibited malnutrition and hyperbilirubinemia.26,27
We observed that female sex was associated with greater morbidity, similar to a previous study from Kyoto that reported that female sex and high GRWR were independent risk factors for hepatic artery thrombosis after LDLT.28
To our knowledge, few studies have reported outcomes of ICU-bound pediatric patients after LDLT.3-5 Hereby, in ICU-bound pediatric patients, bacterial infection was diagnosed in 63% patients, fungal infection was diagnosed in 7.4% patients, and viral infection was diagnosed in 28% patients. Most bacterial and fungal infections occurred in the first 2 weeks after liver transplant. The lower incidence of bacterial infection in our center than previously reported3-5 may be attributed to the lower trough level used in our patients and lower associated immunosuppression. Although the predominant infections in our study were in the abdomen (43%) and bloodstream (26%), similar to other studies, the major isolated pathogens were MRSA, a condition that necessitates special attention with respect to careful drug monitoring to avoid further emergence of drug-resistant organisms.3-5
In conclusion, critically ill children who were admitted to ICU prior to LDLT had worse outcome than other children did. They had marked impairment in liver and renal function and coagulation profile, received more fresh frozen plasma transfusions, and had greater incidence of biliary leakage after surgery. The pretransplant GGTP level had independent prognostic significance for recipient survival. Female sex, pretransplant platelet count, and pretransplant lactate dehydrogenase were associated with poor patient survival. The ICU-bound children had higher infection rates and were complicated with fungal and multidrug resistant bacterial infections.
Volume : 13
Issue : 1
Pages : 100 - 107
DOI : 10.6002/ect.mesot2014.O55
From the 1Department of Anesthesia, Assiut University, Assiut,
Egypt; the 2Department of Anesthesia, King Abdullah Medical City,
Mecca, Saudi Arabia; the 3Department of Surgery, Sohag University,
Sohag, Egypt; the 4Department of Microbiology and Immunology, Assiut
University, Assiut, Egypt; the 5Department of Hematology and
Immunology, Umm Al-Qura University, Mecca, Saudi Arabia; and the 6Department
of Hepato-Pancreato-Biliary Surgery and Transplantation, Kyoto University,
Acknowledgements: Hamed Elgendy, Hanaa Nafady-Hego, Walid El Moghazy, and Shinji Uemoto carried out the research and wrote the manuscript. Hanaa Nafady-Hego, Hamed Elgendy, and Shinji Uemoto participated in research design. Hamed Elgendy, Hanaa Nafady-Hego, and Walid El Moghazy conducted the data analysis. Shinji Uemoto directed the transplant program. All authors have no potential interest to declare. All authors have no funding for this manuscript. Part of this work was selected for oral presentation at the 14th Congress of the Middle East Society for Organ Transplantation (MESOT), September 2014, Istanbul, Turkey.
Corresponding author: Hamed Elgendy, MD, PhD, Department of Anesthesia, King Abdullah Medical City, Post Office Box 57656, Muzdaifah Road, Postal Code 21955, Mecca, Saudi Arabia
Phone: +966 5 0531 6200
Fax: +966 2 553 2239
Table 1. Comparison of Indications for Liver Retransplant in Early and Late Retransplant Groups
Table 2. Intraoperative and Postoperative Factors Involved in Pediatric Recipients Who Underwent Primary Living-Donor Liver Transplant at Kyoto University Hospital, Japan
Table 3. Univariate and Multivariate Analysis of Risk Factors for Survival of Intensive Care Unit-Bound Pediatric Patients Who Underwent Primary Living-Donor Liver Transplant at Kyoto University Hospital, Japan
Table 4A. Site of Infection and Frequency of Isolated Organisms
Table 4B. Site of Infection and Frequency of Isolated Organisms in Hospitalized Patients
Table 4C. Site of Infection and Frequency of Isolated Organisms in At Home Patients
Table 5. Types of Bloodstream Infection
Table 6. Frequency and Rate of Diagnosed Infections
Figure 1. Survival of Pediatric Patients After Living-Donor Liver Transplant
Figure 2. Graft Survival of Pediatric Patients After Living-Donor Liver Transplant