Portal de Periódicos da CAPES

Risk Factors for Rejection and Infection in Pediatric Liver Transplantation

2008; Elsevier BV; Volume: 8; Issue: 2 Linguagem: Inglês

10.1111/j.1600-6143.2007.02068.x

1600-6143

Ross W. Shepherd, Yumirle P. Turmelle, Michelle Nadler, Jeffrey A. Lowell, M R Narkewicz, S V McDiarmid, Ravinder Anand, Chenlong Song,

Clinical Nutrition and Gastroenterology

Rejection and infection are important adverse events after pediatric liver transplantation, not previously subject to concurrent risk analysis. Of 2291 children (<18 years), rejection occurred at least once in 46%, serious bacterial/fungal or viral infections in 52%. Infection caused more deaths than rejection (5.5% vs. 0.6% of patients, p < 0.001). Early rejection (<6 month) did not contribute to mortality or graft failure. Recurrent/chronic rejection was a risk in graft failure, but led to retransplant in only 1.6% of first grafts. Multivariate predictors of bacterial/fungal infection included recipient age (highest in infants), race, donor organ variants, bilirubin, anhepatic time, cyclosporin (vs. tacrolimus) and era of transplant (before 2002 vs. after 2002); serious viral infection predictors included donor organ variants, rejection, Epstein-Barr Virus (EBV) naivety and era; for rejection, predictors included age (lowest in infants), primary diagnosis, donor-recipient blood type mismatch, the use of cyclosporin (vs. tacrolimus), no induction and era. In pediatric liver transplantation, infection risk far exceeds that of rejection, which causes limited harm to the patient or graft, particularly in infants. Aggressive infection control, attention to modifiable factors such as pretransplant nutrition and donor organ options and rigorous age-specific review of the risk/benefit of choice and intensity of immunosuppressive regimes is warranted. Rejection and infection are important adverse events after pediatric liver transplantation, not previously subject to concurrent risk analysis. Of 2291 children (<18 years), rejection occurred at least once in 46%, serious bacterial/fungal or viral infections in 52%. Infection caused more deaths than rejection (5.5% vs. 0.6% of patients, p < 0.001). Early rejection (<6 month) did not contribute to mortality or graft failure. Recurrent/chronic rejection was a risk in graft failure, but led to retransplant in only 1.6% of first grafts. Multivariate predictors of bacterial/fungal infection included recipient age (highest in infants), race, donor organ variants, bilirubin, anhepatic time, cyclosporin (vs. tacrolimus) and era of transplant (before 2002 vs. after 2002); serious viral infection predictors included donor organ variants, rejection, Epstein-Barr Virus (EBV) naivety and era; for rejection, predictors included age (lowest in infants), primary diagnosis, donor-recipient blood type mismatch, the use of cyclosporin (vs. tacrolimus), no induction and era. In pediatric liver transplantation, infection risk far exceeds that of rejection, which causes limited harm to the patient or graft, particularly in infants. Aggressive infection control, attention to modifiable factors such as pretransplant nutrition and donor organ options and rigorous age-specific review of the risk/benefit of choice and intensity of immunosuppressive regimes is warranted. In pediatric liver transplantation, improved patient and graft survival is attributed to advances in surgery and improved immunosuppression regimens (1Colombani PM Dunn SP Harmon WE Magee JC McDiarmid SV Spray TL Pediatric transplantation.Am J Transplant. 2003; 3: 53-63Crossref PubMed Scopus (46) Google Scholar). Advances in immunosuppression have been concentrated on preventing rejection, and indeed the use of such agents has led to a decreased rate of acute cellular rejection and graft loss from rejection (2Martin SR Atkison P Anand R Lindblad AS SPLIT Research GroupStudies of Pediatric Liver Transplantation 2002: Patient and graft survival and rejection in pediatric recipients of a first liver transplant in the United States and Canada.Pediatr Transplant. 2004; 8: 273-283Crossref PubMed Scopus (140) Google Scholar). However, potent immunosuppression imparts a risk for life-threatening infection and other adverse effects. One challenge for the physician managing children after liver transplant is to balance these risks. Outcomes might be improved if modifiable risks are better characterized. Unfortunately, no large-scale studies have concurrently evaluated outcomes and risk factors for both rejection and all types of infection. The imperative for such a study is emphasized by recent data, which suggest that morbidity and mortality from infections in children after liver transplantation may exceed those for rejection (3Utterson EC Shepherd RW Sokol RJ et al.Biliary atresia: Clinical profiles, risk factors, and outcomes of 755 patients listed for liver transplantation.J Pediatr. 2005; 147: 180-185Abstract Full Text Full Text PDF PubMed Scopus (163) Google Scholar,4Jain A Mazariegos G Kashyap R et al.Pediatric liver transplantation: A single center experience spanning 20 years.Transplantation. 2002; 73: 941-947Crossref PubMed Scopus (116) Google Scholar). All children receiving their first transplant enrolled in the SPLIT Registry between 1995 and 2006 were included in this data analysis. All SPLIT centers have Institutional Review Board approval and individual informed consent is obtained from parents and/or guardians (5Hollenbeak CS Alfrey EJ Sheridan K Burger TL Dillon PW Surgical site infections following pediatric liver transplantation: Risks and costs.Transpl Infect Dis. 2003; 5: 72-78Crossref PubMed Scopus (43) Google Scholar, 6Quiros-Tejeira RE Ament ME McDiarmid SV et al.Late-onset bacteremia in uncomplicated pediatric liver-transplant recipients after a febrile episode.Transpl Int. 2002; 15 (Epub 2002 Sep 19.): 502-507Crossref Scopus (2) Google Scholar, 7Their M Holmberg C Lautenschlager I Hockerstedt K Jalanko H Infections in pediatric kidney and liver transplant patients after perioperative hospitalization.Transplantation. 2000; 69: 1617-1623Crossref PubMed Scopus (30) Google Scholar). Coded information is submitted to the SPLIT data-coordination center at the time of listing for liver transplant (LT). Follow-up data were submitted on a biannual basis pre- and post-LT in the first 2 years and yearly, thereafter. There is long-term reporting of data related to events such as LT, death, allograft rejection, posttransplant complications, including infections and serious viral infection [symptomatic Epstein-Barr Virus (EBV) and cytomegalovirus (CMV) disease and EBV-related posttransplant lymphoproliferative disease (PTLD)]. For this study, database elements pertaining to deaths and graft losses due to infection and rejection were analyzed and causes of death and graft failure tabulated, including probability of posttransplant patient and graft survival, probability of rejection, probability of bacterial and fungal infection in the first 30 days and probability of serious viral infection in the first 15 months posttransplant. These time periods were chosen for the purposes of the detailed risk analyses for infections, as they represent the peak periods for the development of these types of infections (5Hollenbeak CS Alfrey EJ Sheridan K Burger TL Dillon PW Surgical site infections following pediatric liver transplantation: Risks and costs.Transpl Infect Dis. 2003; 5: 72-78Crossref PubMed Scopus (43) Google Scholar, 6Quiros-Tejeira RE Ament ME McDiarmid SV et al.Late-onset bacteremia in uncomplicated pediatric liver-transplant recipients after a febrile episode.Transpl Int. 2002; 15 (Epub 2002 Sep 19.): 502-507Crossref Scopus (2) Google Scholar, 7Their M Holmberg C Lautenschlager I Hockerstedt K Jalanko H Infections in pediatric kidney and liver transplant patients after perioperative hospitalization.Transplantation. 2000; 69: 1617-1623Crossref PubMed Scopus (30) Google Scholar, 8George DL Arnow PM Fox A et al.Patterns of infection after pediatric liver transplantation.Am J Dis Child. 1992; 146: 924-929PubMed Google Scholar, 9Sokal EM Antunes H Beguin C et al.Early signs and risk factors for the increased incidence of Epstein-Barr virus-related posttransplant lymphoproliferative diseases in pediatric liver transplant recipients treated with tacrolimus.Transplantation. 1997; 64: 1438-1442Crossref PubMed Scopus (161) Google Scholar, 10Mellon A Shepherd RW Faoagali JL et al.Cytomegalovirus infection after liver transplantation in children.J Gastroenterol Hepatol. 1993; 8: 540-544Crossref PubMed Scopus (24) Google Scholar). Infants are defined as 16-fold). Open table in a new tab Primary causes of death after first and subsequent transplants in 2291 children as listed in the SPLIT database. Expressed as number (n) and percentage (%) of total deaths after each transplant. Note that infection caused significantly more deaths than rejection (>16-fold). Table 1 tabulates the primary cause of death following first or subsequent transplants. Infection was the most common cause of death, listed as the primary cause of death in 70 patients (3.1% of all patients), the majority being due to bacterial sepsis (56%), with viral and fungal infections accounting for 19% and 10% of infection deaths, respectively. Infection accounted for 25% of first transplant deaths, 27% of second transplant deaths and 30% of third transplant deaths. Infection was also listed as contributing to death in another 74 patients whose primary cause of death was either multi-organ failure, cardiopulmonary failure or less commonly graft (liver) failure (data not shown). Infection therefore directly or indirectly contributed to the deaths of 125/2291 (5.5%) of patients, accounting for 125/274 (46%) of deaths. In contrast, rejection directly or indirectly contributed to the deaths of 13/2291 (0.6%) of all patients, accounting for 13/274 (4.7%) of deaths. Rejection was the primary cause of death via graft failure in only 4 patients overall (0.2% of all patients, 1.5% of deaths), although rejection may have contributed to the death of another 9 patients, because rejection was the reason for retransplant in 36 patients, 9 of whom died after retransplant (see below). By univariate analysis, rejection was not a risk factor in mortality, including rejection in the first 6 months after transplant, and or comparison of 0 versus 1, or >1 versus 0 episodes of rejection overall (respective hazards ratios 0.986, 0.873 and 1.918, p-values = 0.94, 0.52 and 0.07, respectively). Thus overall, there was a 10-fold greater risk of death from infection than from rejection (3.1% vs. 0.2% of patients or 46% vs. 4.7% of deaths, p < 0.001). Other causes of death (Table 1) included multiorgan failure and cardiopulmonary failure (n = 127, 46% of all deaths), graft failure (either first or second graft) without the benefit of further retransplant (n = 50), brain injury (n = 27) and secondary malignancy (n = 14). As shown in Table 2, 236/2291 (10.3%) of patients receiving a primary graft had retransplants for graft failure, with vascular or surgical complications and primary graft nonfunction accounting for the majority (36% and 23%, respectively). Two retransplants were directly attributable to infection in the graft (hepatitis C). Rejection was a reason for graft failure and retransplant in 36/236 (15%) of the first retransplants, three-fourths of these for chronic rejection. It follows that most of the rejection was treatable, with only 1.6% of all patients requiring retransplant for graft failure due to rejection.Table 2Causes of Graft Failure leading to Retransplantation after first and subsequent transplants in 2291 children as listed in the SPLIT databaseTransplants (number)1234Patients2291236322Retransplanted2363221Primary reason for retransplantPrimary graft dysfunction545Hyperacute rejection2Acute rejection81Chronic rejection26Ductopenic211Vascular5Vascular/postoperative complication8414Hepatic artery thrombosis6591Portal vein thrombosis1751Postoperative hemorrhage2Biliary tract complication14Intrahepatic only5Intra and extrahepatic91InfectionHCV infection2Poor compliance2Recurrent liver disease5Other177Note that of 236 second transplants, rejection accounted for only 36/236 (15%) of the primary graft failures. There are 22/236 patients with missing primary reasons for the second transplant and 4/32 with missing primary reasons for the third transplant. Open table in a new tab Note that of 236 second transplants, rejection accounted for only 36/236 (15%) of the primary graft failures. There are 22/236 patients with missing primary reasons for the second transplant and 4/32 with missing primary reasons for the third transplant. By univariate anlaysis, neither the occurrence of rejection episodes in the first 6 months after transplant or a comparison of 0 versus 1 episode of rejection predicted graft survival (respective hazards ratio 1.205, p = 0.24 and 1.096, p = 0.59). Comparing 0 and >1 episode of rejection, the hazards ratio, was 2.440, p = 0.006. In those patients receiving two or more retransplants for any reason, only two patients had further graft failures due to rejection and one of them had previous retransplantation due to rejection. Thus, rejection in the first 6 months and single episodes of rejection did not contribute to graft failure, although recurrent rejection was a risk factor. Overall, rejection contributed to graft failure and retransplantation in only 36/2291 (1.6%) of all patients. About 45% of patients developed at least one episode of rejection within 6 months of transplant, 38% developed serious bacterial or fungal infections (<30 days) and 14% had serious viral infections <15 months after liver transplant. There was a significant age-related variance in both infection and rejection rates. The rejection rate was 2-fold lower in those transplanted as infants compared with adolescents (0.20 vs. 0.44 episodes per patient year, p < 0.001), although the mean time to first rejection episode (156 ± 43 days vs. 103 ± 16 days posttransplant) was not statistically significant (p = 0.40). Conversely, the bacterial or fungal infection rate was highest in infants and lowest in adolescents (50% vs. 21%, p < 0.001), as was the viral infection rate (15% vs. 11%, p = 0.06). Of the 762 bacterial infections documented, 39% were line infections, 35% were intra-abdominal, 18% were bacterial sepsis, 14% were wound infections, 17% were urinary infections, 13% were pneumonia and 7% were cholangitis. Of the 189 fungal infections, 32% were intra-abdominal, 27% were urinary, 14% were line and 12% were lung infections. For serious viral infections, the overall CMV disease rate was 6% and the EBV disease rate was 8.6%, with a PTLD rate of 2.7%. About half those with CMV disease more common in the first 30 days (3%) developed this within the first 30 days, whereas most (90%) of those with EBV disease (8%) occurred beyond 30 days posttransplant. Adenovirus caused pneumonia or gastrointestinal illness in about 1% (n = 22). On univariate analysis, age was a significant risk factor in the development of CMV or EBV disease (data not shown), and the PTLD rate was 10-fold higher in patients transplanted as infants compared to that in adolescents (3.5% vs. 0.3%, p < 0.001). CMV and EBV serological status at the time of transplant was recorded in 2226 cases for CMV and 2157 cases for EBV. Of these, 64% were CMV-naive and 63% were EBV-naive children at the time of transplant. Of those 1 implies patients in category A have higher risk of outcome compared with category B. Relative risks and the corresponding confidence intervals are adjusted for other factors in the model.p-Value95% CI2CI = Confidence intervals.Recipient’s age6–11 month0–5 month1.390.032(1.03, 1.87)1–4 years1.84<.0001(1.39, 2.45)5–12 years1.570.003(1.16, 2.12)≥13 years1.93<.0001(1.39, 2.67)Primary diagnosisOther cholestatic or metabolicBiliary atresia0.830.037(0.69, 0.99)Fulminant liver failure1.050.699(0.83, 1.31)Cirrhosis0.890.408(0.67, 1.18)Other0.750.040(0.57, 0.99)Donor-recipient blood matchCompatibleIdentical0.920.430(0.75, 1.13)Incompatible0.550.028(0.32, 0.94)First immunosuppressionCyclosporineTacrolimus1.49<.0001(1.27, 1.74)IV IG use within the first week post-TxYesNo0.750.002(0.62, 0.89)Year of transplant≥2002≤20010.69<.0001(0.60, 0.81)B. Bacterial/ fungal infectionFactorCategoryOutcome infectionAB (Reference)Odds Ratiop-Value95% CI2Recipient’s age6–11 months0–5 months1.4500.1066(0.923, 2.276)1–4 years1.0520.8221(0.674, 1.644)5–12 years0.7010.1510(0.432, 1.138)≥13 Years0.4490.0053(0.256, 0.788)RaceBlackWhite1.4080.0553(0.992, 1.999)Hispanic1.6190.0035(1.172, 2.237)Other1.4230.1422(0.888, 2.278)Donor/organ typeCad-reducedWhole1.7440.0014(1.239, 2.454)Cad-Split2.461<.0001(1.643, 3.686)Live1.3880.0885(0.952, 2.024)First immunosuppressionCyclosporineTacrolimus1.3960.0419(1.012, 1.925)Year of transplant≥2002≤20010.7030.0100(0.538, 0.919)Log total bilirubinContinuous predictor1.1330.0213(1.019, 1.260)Log anhepatic timeContinuous predictor1.4090.0153(1.068, 1.860)C. Viral infectionsFactorCategoryOutcome infectionAB (Reference)RRp-Value95% CI2Donor/organ typeCad-reducedWhole1.8650.0005(1.254, 2.264)Cad-split1.0310.8899(0.671, 1,584)Live1.1880.3293(0.840, 1.680)Year of transplant≥2002≤20010.7260.0134(0.563, 0.936)Rejection Post-TxYesNo1.705 0.05).1 Relative risk >1 implies patients in category A have higher risk of outcome compared with category B. Relative risks and the corresponding confidence intervals are adjusted for other factors in the model.2 CI = Confidence intervals. Open table in a new tab Rejection is defined as treated episodes, bacterial or fungal infections as culture-proven infections in the first 30 days, and serious viral infections as cytomegalovirus, Epstein-Barr virus disease or lymphoproliferative disease. Of 22 risk factors evaluated by univariate analysis, those significant at a p-value of 0.1 for bacterial, fungal or viral infections and 0.15 for rejection were used in the multivariate model, and a backward-elimination procedure was performed to obtain those significant risk factors depicted here (p-value >0.05). For bacterial infections, significant factors in the univariate analysis included a wide range of demographic factors (age, race, primary diagnosis and era of transplant), severity factors (height and weight deficit, PELD score and its components, bilirubin, albumin and WBC levels at transplant, time on the waiting list, patient acuity status), surgical factors (prior abdominal surgery, donor organ type, anhepatic time) and immunological factors (rejection, donor age and cyclosporine vs. tacrolimus at initiation). In the multivariate analysis (Table 3, Part B), age, race, immunosuppression, year of transplant, bilirubin level and organ donor type were significant independent risk factors. Of note, infants had higher odds ratio for bacterial infections than adolescents. Those who received deceased donor split or reduced-sized liver also were at higher risk. The latter are more likely to be transplants in infants, thereby heightening this risk in this age group. For viral infections, in the univariate analysis, significant risk factors included age, and were predominantly immunological, including rejection, cyclosporin use and era of transplant. Various disease severity factors were not significant, but as for bacterial infections, those who received deceased donor split or reduced-sized liver also were at higher risk. Of these significant factors in the univariate analysis, only rejection (relative risk 1.65), era of transplant and organ donor variants were significant in the multivariate analysis (Table 3, Part C). This analysis of data, derived from the largest cumulative dataset of pediatric liver transplants available, describes outcomes and risk factors in relation to rejection and infection, both important and potentially inter-related adverse events after liver transplantation. There is no similar concurrent analysis with which to compare these data, which provide a broad view of outcomes across centers in North America, although earlier less focussed analyses from the same database stimulated this concurrent study (2Martin SR Atkison P Anand R Lindblad AS SPLIT Research GroupStudies of Pediatric Liver Transplantation 2002: Patient and graft survival and rejection in pediatric recipients of a first liver transplant in the United States and Canada.Pediatr Transplant. 2004; 8: 273-283Crossref PubMed Scopus (140) Google Scholar,3Utterson EC Shepherd RW Sokol RJ et al.Biliary atresia: Clinical profiles, risk factors, and outcomes of 755 patients listed for liver transplantation.J Pediatr. 2005; 147: 180-185Abstract Full Text Full Text PDF PubMed Scopus (163) Google Scholar,11McDiarmid SV Anand R SPLIT Research GroupStudies of Pediatric Liver Transplantation (SPLIT): A summary of the 2003 Annual Report.Clin Transpl. 2003; : 119-130Google Scholar). Both infection and rejection are relatively common occurrences, but these data confirm that the risk from infection far exceeds that from rejection, particularly in infants, a disparity that deserves detailed analysis. Rejection rarely contributed to and was not a risk factor in mortality, and the risk of graft failure from rejection was low, limited to chronic or recurrent rejection, which in itself was uncommon. Single episodes of rejection and rejection in the first 6 months were not predictors of graft failure, suggesting that acute cellular rejection was almost always treatable. In contrast, infection was the most common cause of death and clearly caused much more morbidity than rejection. Young age was an important risk factor for infections, but a negative risk factor for rejection, and infants had half the rate of rejection, three times the rate of bacterial or fungal infection and 10 times the rate of PTLD compared with adolescents. These overall risk analysis findings were independent of the relative improvements in the risks of both rejection and infection when comparing before 2002 versus after 2002 (risk ratio = 0.7), which may be explained by improved immunosuppression choices and dose monitoring, and the trend to early steroid withdrawal and immunosuppression minimization (12McDiarmid SV Anand R Lindblad AS SPLIT Research GroupStudies of Pediatric Liver Transplantation: 2002 update. An overview of demographics, indications, timing, and immunosuppressive practices in pediatric liver transplantation in the United States and Canada.Pediatr Transplant. 2004; 8: 284-294Crossref PubMed Scopus (82) Google Scholar). Other risk factors for bacterial and fungal infections included recipient severity of illness factors and surgical issues; and for serious viral infections, the primary immunosuppression used and rejection (presumably via increased immunosuppression). Collectively, these data raise the possibility that choice and/or intensity of primary immunosuppression are modifiable risk factors for infection, particularly for serious viral infections, and most particularly for infants. Certainly, avoiding over-immunosuppression by carefully monitoring calcinuerin inhibitor dosage and blood levels and limiting steroid use at least in this age group would be pertinent goals. Infants are innately in a state of immune immaturity, are more likely CMV and EBV naive, but pose much less of a threat for organ failure due to rejection, and could benefit from immunosuppression minimization. Thus, a rigorous age-specific review of the choice and intensity of immunosuppression, balanced against the relatively low risk of rejection is warranted in pediatric liver transplantation. Of relevance to these data is the development and increasing use of newer, often more potent immunosuppressive regimens (12McDiarmid SV Anand R Lindblad AS SPLIT Research GroupStudies of Pediatric Liver Transplantation: 2002 update. An overview of demographics, indications, timing, and immunosuppressive practices in pediatric liver transplantation in the United States and Canada.Pediatr Transplant. 2004; 8: 284-294Crossref PubMed Scopus (82) Google Scholar, 13McDiarmid SV Anand R Lindblad AS Principal Investigators and Institutions of the Studies of Pediatric Liver Transplantation (SPLIT) Research GroupDevelopment of a pediatric end-stage liver disease score to predict poor outcome in children awaiting liver transplantation.Transplantation. 2002; 74: 173-181Crossref PubMed Scopus (301) Google Scholar, 14Al-Hussaini A Tredger JM Dhawan A Immunosuppression in pediatric liver and intestinal transplantation: A closer look at the arsenal.J Pediatr Gastroenterol Nutr. 2005; 41: 152-165Crossref Scopus (14) Google Scholar, 15Schuller S Wiederkehr JC Coelho-Lemos IM Avilla SG Schultz C Daclizumab induction therapy associated with tacrolimus-MMF has better outcome compared with tacrolimus-MMF alone in pediatric living donor liver transplantation.Transplant Proc. 2005; 37: 1151-1152Crossref PubMed Scopus (22) Google Scholar, 16Heffron TG Pillen T Smallwood GA Welch D Oakley B Romero R Pediatric liver transplantation with daclizumab induction.Transplantation. 2003; 75: 2040-2043Crossref PubMed Scopus (49) Google Scholar, 17Ganschow R Grabhorn E Schulz A Von Hugo A Rogiers X Burdelski M Long-term results of basiliximab induction immunosuppression in pediatric liver transplant recipients.Pediatr Transplant. 2005; 9: 741-745Crossref PubMed Scopus (41) Google Scholar, 18Tredger JM Brown NW Dhawan A Immunosuppression in pediatric solid organ transplantation: opportunities, risks, and management.Pediatr Transplant. 2006; 10: 879-892Crossref PubMed Scopus (31) Google Scholar), for example induction agents, poly- or monoclonal antibodies, mycophenylate and sirolimus. It is unfortunate that most studies that use these newer agents often initially emphasize attaining an almost zero rejection rate, but have not included a detailed evaluation of infection risk (15Schuller S Wiederkehr JC Coelho-Lemos IM Avilla SG Schultz C Daclizumab induction therapy associated with tacrolimus-MMF has better outcome compared with tacrolimus-MMF alone in pediatric living donor liver transplantation.Transplant Proc. 2005; 37: 1151-1152Crossref PubMed Scopus (22) Google Scholar, 16Heffron TG Pillen T Smallwood GA Welch D Oakley B Romero R Pediatric liver transplantation with daclizumab induction.Transplantation. 2003; 75: 2040-2043Crossref PubMed Scopus (49) Google Scholar, 17Ganschow R Grabhorn E Schulz A Von Hugo A Rogiers X Burdelski M Long-term results of basiliximab induction immunosuppression in pediatric liver transplant recipients.Pediatr Transplant. 2005; 9: 741-745Crossref PubMed Scopus (41) Google Scholar). For example, substitution of CNIs by mTOR inhibitors such as sirolimus may be promising as a substitute for patients with calcineurin inhibitor nephrotoxicity, but their use requires validation in long-term studies in large cohorts, particularly with regard to the increasingly reported risk of serious interstitial pneumonia and other infections (18Tredger JM Brown NW Dhawan A Immunosuppression in pediatric solid organ transplantation: opportunities, risks, and management.Pediatr Transplant. 2006; 10: 879-892Crossref PubMed Scopus (31) Google Scholar,19Champion L Stern M Israel-Biet D et al.Brief communication: Sirolimus-associated pneumonitis: 24 cases in transplant recipients.Ann Intern Med. 2006; 144: 505-509Crossref PubMed Scopus (169) Google Scholar). In addition, age responsiveness and risk seem critical when evaluating the use of new agents. For example, the use of mycophenylate or induction agents may be unnecessary in infants given their very low risk of rejection and more risky because of the innate immaturity of their immune system. In all age groups however, potent immunosuppressive regimens have other drawbacks (18Tredger JM Brown NW Dhawan A Immunosuppression in pediatric solid organ transplantation: opportunities, risks, and management.Pediatr Transplant. 2006; 10: 879-892Crossref PubMed Scopus (31) Google Scholar,20Bucuvalas JC Campbell KM Cole CR Guthery SL Outcomes after liver transplantation: keep the end in mind.J Pediatr Gastroenterol Nutr. 2006; 43: S41-S48Crossref PubMed Scopus (24) Google Scholar) besides risk for infection, including malignancy, metabolic adverse effects, growth failure and late renal insufficiency. A key question thus is, what level of rejection is acceptable? These data emphasize the concept of rejection risk (to the patient and graft) rather than rejection rate. While at least one episode of rejection occurred in about half the patients, the overall rejection rate was 0.29 episodes per patient year (0.20 in infants and 0.44 in adolescents), and only 1.5% of patients developed graft failure, with few patients dying directly or indirectly due to rejection. The era effect does indicate that rejection rates are now acceptably low and chronic rejection is almost absent in the tacrolimus era, confirming recent large single center analyses (21Jain A Mazariegos G Pokharna R et al.The absence of chronic rejection in pediatric primary liver transplant patients who are maintained on tacrolimus-based immunosuppression: A long-term analysis.Transplantation. 2003; 75: 1020-1025Crossref PubMed Scopus (35) Google Scholar). Moreover, recent evidence suggests that the immune response toward a liver allograft is not necessarily always harmful. Some alloreactivity may actually facilitate graft tolerance, which clearly does occur in a significant number of patients (22Mazariegos GV Sindhi R Thomson AW Marcos A Clinical tolerance following liver transplantation: Long term results and future prospects.Transpl Immunol. 2007; 17 (Epub 2006, Oct 17.): 114-119Crossref PubMed Scopus (86) Google Scholar,23Koshiba T Li Y Takemura M et al.Clinical, immunological, and pathological aspects of operational tolerance after pediatric living-donor liver transplantation.Transpl Immunol. 2007; 17 (Epub 2006, Nov 10): 94-97Crossref PubMed Scopus (167) Google Scholar), explained variously by the opposing theories of clonal exhaustion-deletion (24Starzl TE Zinkernagel RM Transplantation tolerance from a historical perspective.Nat Rev Immunol. 2001; 1: 233-239Crossref PubMed Scopus (164) Google Scholar) or ‘allo-immune homeostasis’ (25Miller J Mathew JM Esquenazi V Toward tolerance to human organ transplants: A few additional corollaries and questions.Transplantation. 2004; 77: 940-942Crossref PubMed Scopus (12) Google Scholar). Infants are more ‘tolerogenic’, not entirely explained by these theories, as indicated by the clinical tolerance of ABO-incompatible organs in infant heart transplant recipients (26Daebritz SH Schmoeckel M Mair H Kozlik-Feldmann R Wittmann G Kowalski C Kaczmarek I Reichart B Blood type incompatible cardiac transplantation in young infants.Eur J Cardiothorac Surg. 2007; 31 (discussion 343. (Epub 2007, Jan 17.)): 339-343Crossref PubMed Scopus (25) Google Scholar). A large study of outcomes from acute rejection in adult liver transplantation, where some rejection actually conveyed a graft outcome advantage, raised the question as to whether complete elimination of all rejection is really a desirable goal in liver transplantation (27Wiesner RH Demetris AJ Belle SH et al.Acute hepatic allograft rejection: Incidence, risk factors, and impact on outcome.Hepatology. 1998; 28: 638-645Crossref PubMed Scopus (408) Google Scholar). While there is some room for improvement in recurrent rejection as a risk factor for some graft loss, based on the graft failure data presented here, this has to be tempered by the overall risk to the patient. What may be required to make this improvement is not more potent immunosuppression but a more specific and perhaps ‘smarter individualized approach’. Questions such as, do infants need steroids, mycophenylate or induction therapy at all? How intensely should pediatric liver recipients be immunosuppressed? And in which recipients can immunosuppression be discontinued altogether, all require study. It is probably time to submit these to rigorous controlled trials, rather than submit to the trial and error approach of the past. Nonetheless, the data presented here give some confidence to the current emphasis toward immunosuppression minimization in pediatric liver recipients. Other factors, including improvements in the ability to treat perioperative infection, monitoring for and preventing viral infections, and improved focus on transplantation when the child is in better condition are also emphasized as likely to be leading to improved outcomes and reducing deaths from infections. An individualized approach to the prevention of EBV disease and PTLD (28Lee TC Goss JA Rooney CM et al.Quantification of a low cellular immune response to aid in identification of pediatric liver transplant recipients at high-risk for EBV infection.Clin Transplant. 2006; 20: 689-694Crossref PubMed Scopus (48) Google Scholar), would be particularly relevant to infants, given the high risk of these problems in CMV- or EBV-naive patients. Undernutrition is a well-documented risk factor, particularly in biliary atresia (3Utterson EC Shepherd RW Sokol RJ et al.Biliary atresia: Clinical profiles, risk factors, and outcomes of 755 patients listed for liver transplantation.J Pediatr. 2005; 147: 180-185Abstract Full Text Full Text PDF PubMed Scopus (163) Google Scholar), modifiable by aggressive nutritional support (29Chin SE Shepherd RW Cleghorn GJ et al.Pre-operative nutritional support in children with end-stage liver disease accepted for liver transplantation: An approach to management.J Gastroenterol Hepatol. 1990; 5: 566-572Crossref PubMed Scopus (40) Google Scholar,30Chin SE Shepherd RW Thomas BJ et al.Nutritional support in children with end-stage liver disease: A randomized crossover trial of a branched-chain amino acid supplement.Am J Clin Nutr. 1992; 56: 158-163Crossref PubMed Scopus (128) Google Scholar). For reasons that are not immediately obvious, the type of donor organ (split, deceased donor cut-down and living donor organ vs. whole) appears to be a factor in both bacterial and viral infections, but not rejection. Surgical complications from the use of these donors may increase the potential for bacterial and fungal infections. In conclusion, in pediatric liver transplantation, infection risk far exceeds that of rejection, which on current immunosuppression regimes causes limited harm to the patient or graft. Rigorous evaluation of the choice and intensity and monitoring of immunosuppression regimens in pediatric liver transplantation is indicated, especially in infant recipients, where infection risk is highest and rejection risk is lowest. The current trend toward immunosuppression minimization is supported by these data. However, anticipatory management and aggressive control of infection, and attention to modifiable factors such as pretransplant nutrition, and appropriate choice of donor organ options seems advisable. Any new immunosuppressive regime requires age-specific evaluation, and concurrent analysis of infection risk. Findings from this study may help in decision making in the choice of immunosuppressive regimens and call attention to the concept of reducing rejection risks rather than rates as a preferable goal in pediatric liver transplantation, balancing these risks against the risks of immunosuppression. SPLIT is supported by NIDDK Grant #U01 DK061693–01A1 and an Educational Grant from Astellas. Dr Turmelle was a 2006–2007 Advanced Trainee in Hepatology supported by the American Liver Foundation.