Although single-center reports regarding LGL and mortality have been published in the past, these studies were focused on mixed adult and pediatric patient populations and were limited by the relatively small number of patients experiencing LGL or LM (3–8). The size and homogeneity of data in the SPLIT database overcomes these limitations. The database also allows for a more comprehensive and current look at late mortality and graft loss using data acquired from multiple centers. Several studies have utilized the 1-year time point to define “late” events (3–5). The majority of graft losses after LT occur early in the postoperative period, usually in the first few months. Graft failure in this time is mostly due to primary nonfunction and thrombosis with early mortality due to infection, PTLD and bleeding (10). In clinical practice, most of these complications are resolved by the first posttransplant year. In attempting to better understand late events after LT and the complex interaction between factors associated with late outcomes after LT, analysis is best accomplished looking at LGL and LM separately.
Late graft loss
The SPLIT database demonstrated that the majority of LGLs were caused by acute and chronic rejection (48.5% of all LGL) and by the chronic effects of hepatic arterial thrombosis (11.4% of LGL) and biliary strictures (8.6% of LGL).
In a review from Belgium, Wallot et al. described the etiology and risk factors for graft loss occurring after the third postoperative month (8). Although the study included children transplanted across all eras of transplant and with varied immunosuppressive regimens, a striking number of graft losses were found to be due to immunosuppression reduction associated with PTLD. Similar to our study, biliary and arterial complications as well as chronic rejection (17%) remain a significant contributor to LGL. The concerning impact of chronic rejection on LGL seen in this study is confirmed by our findings. Although small in absolute numbers, there is clearly a subset of patients who suffer LGL due to chronic rejection, contrasting published accounts of an almost insignificant incidence of this complication (9,10).
Those patients who experience more than one episode of acute cellular rejection were found to have an almost two-fold increase in LGL, a finding that reaffirms recent studies that demonstrate an association between chronic rejection and recurrent acute cellular rejections (8,11). Similar to other studies, steroid-resistant rejection was also found to be associated with LGL (8). Although tacrolimus has been shown to dramatically decrease the rate of chronic rejection in pediatric liver-transplant recipients (9–12), analysis of the SPLIT data did not show any difference in LGL in those children receiving cyclosporine versus tacrolimus (data not shown). Our data also demonstrated that the incidence of LGL was similar in patients transplanted before 2002 when compared to those receiving LT after 2002.
Our analysis also shows that patients who required multiple operations and those requiring frequent readmission during the first posttransplant year have an increased incidence of LGL, suggesting that early technical complications and infections arising from the initial transplant procedure significantly impair long-term graft-survival. It is also important to note that unlike earlier studies, technical variant grafts (living or deceased donor) had similar rates of LGL when compared to whole allografts (1,6,8,10).
The solitary graft loss in patients transplanted for tumor was due to chronic rejection. Although multivariate analysis demonstrated an association between transplant for tumor and LGL, small sample size makes interpretation of these results difficult.
Death in children who survive their first posttransplant year is caused by sepsis, multisystem organ failure and posttransplant lymphoproliferative disease. In contrast to the adult transplant population (13,14), recurrent disease and cardiopulmonary diseases are not common findings after pediatric LT. Approximately 30% of the SPLIT patients who died after their first posttransplant year succumbed due to infection, sepsis or MSOF. Although the data is limited by the relative lack of details surrounding these deaths, mortality due to sepsis is much more common after LT compared to the general population. A recent single center review found 60% of LM occurring due to septic complications and PTLD with the remainder of deaths caused by recurrent malignancy (3).
Similar to the findings in LGL analysis, frequent readmission in the first year is associated with late mortality. Although HAT and biliary complications contribute to LM, only three patients had significant liver dysfunction as the primary cause of LM. The large majority of late deaths in this study were thus due to malignancy and primary infections not associated with technical issues.
Children who were 0–2 SD below the mean for weight were also more likely to suffer from LM when compared to those above the mean. The effect of malnutrition on pre- and posttransplant outcomes has been clearly demonstrated in the past, with dramatically reduced posttransplant survival if transplanted with Z scores for weight of <1.0 (15). The reason for this remains unclear, though severe preoperative malnutrition associated with liver disease may play a role in the recipient's ability to overcome the metabolic stresses associated with liver transplantation.
In our study, the majority of the 34 children who died had weight z scores at transplant below the mean as compared to 252/838 (31%) children with weight z scores above the mean who survived. A total of 113 children had significant weight deficit at transplant that continued to year 1. Seven of these patients subsequently died from sepsis (n = 1), PTLD (n = 2), multiple organ failure (n = 1), cardiopulmonary causes (n = 2) and metastatic liver malignancy in one child. Interestingly, patients who were greater than 2 SD below the mean in weight at the time of transplant did not have an increased risk of LM. Although one can only speculate about the reasons behind this finding, it is possible that the weight in some of these patients is artificially elevated due to ascites or that a larger proportion of patients transplanted with this weight deficit experienced mortality early in the course of LT. This issue could be resolved by a more specific examination of the impact of perioperative nutrition and height and weight at the time of transplant in the SPLIT population.
EBV-related disease remains an important cause of LM after LT. Fortunately, most large centers have adopted monitoring and preemptive therapy protocols similar to those published by UCLA and Pittsburgh (16,17). Notably, both series enjoyed dramatic decreases in the incidence and mortality of PTLD since instituting monitoring and preemptive therapy strategies based on serial measurements of quantitative EBV DNA by PCR (3,18). These strategies have also been successful in decreasing the incidence of severe perioperative CMV infections. The prevention of EBV-related tumors and the introduction of successful treatments for PTLD likely account for the lack of any correlation between symptomatic CMV and EBV infections, PTLD and either LGL or LM in this study. The fact that the development of these complications was not associated with either late outcome in our study suggests that EBV-related complications are being successfully treated without causing LGL due to acute or chronic rejection. This represents an important advance, which is likely due to the development of specific therapies for PTLD and highlights the importance of immune monitoring in this unique group of patients.
Recurrent malignancy remains the Achilles' heel of transplantation for nonresectable childhood hepatic tumors. Patients receiving LT for cancer suffered from a near ten-fold risk of late mortality when compared to a standard group transplanted for biliary atresia. These results differ from those of earlier analyses in which pretransplant diagnosis was not associated with late outcome. Of note, preoperative variables, such as tumor size, type and stage were not investigated in this study and we thus suggest caution in the selection of patients receiving a LT for tumor. In addition, patients transplanted for tumor often receive perioperative cytotoxic chemotherapy, which may ultimately increase their risk of serious infections and the development of other denovo malignancies. Two-thirds of LM in this study group was due to recurrence and metastases of the tumor, with the remaining mortality due to cardiomyopathy, likely a complication of chemotherapy. These results also support the establishment of registry efforts to document all transplant variables in pediatric patients undergoing transplant for liver tumors and to develop prospective analysis of outcomes and optimal timing of transplantation.
Fulminant hepatic failure is one of the most challenging clinical situations in pediatric transplantation. As these children do not suffer from chronic illness, one would expect excellent long-term results with survival past the perioperative period. However, this has not proven to be the case. Early mortality in patients transplanted for FHF is higher than that after LT for other conditions (19,20). Our study is the first to demonstrate that even a year after successful LT, patients who suffered from FHF have a markedly increased risk of dying. Thirty percent of the LM observed in the patients transplanted for FHF in this study was due to multi-organ failure and sepsis. Although the exact etiology of all LM in this population has not been examined fully, a recent review of the SPLIT experience with FHF does suggest increased infectious and neurologic complications after LT for FHF (19). One explanation may be that patients who suffer from FHF have a dysregulation of the natural killer cell lineage, which results in a deranged response to self and other antigens. This dysregulation may also lead to an increased susceptibility to infectious complications after transplantation, which is an important etiology of late death after LT (21).
In summary, LGL in this large multicenter series is related to immunologic factors, rejection and delayed sequelae of technical complications. Despite the traditional perception of the tolerogenicity of the liver and the reported low incidence of chronic rejection currently in liver transplantation, our data suggests that improvement in long-term immunosuppression remains to be achieved. Furthermore, septic complications and malignancy, two factors traditionally associated with over immunosuppression continue to account for the majority of late mortalities.
Although this report is constrained by the natural limitations of database registries, it is the first to examine LGL and mortality in a large multi-center pediatric patient population. Further subgroup analysis is needed to further maximize the impact of these findings to individual patients in unique clinical scenarios. This report underscores the urgent need for individualized patient specific immunosuppression and monitoring that can guide optimal immunosuppression over the long term.
SPLIT Research Group
Stephen Dunn, MD, Jerome Manendez, MSN, RN, Louise Flynn, MSN, RN, Alfred I Dupont Hospital for Children, Wilmington, DE; Maureen Jonas, MD, Laura Krawczuk, CPNP, Marielle Christoff, Boston Children's Hospital, Boston, MA; Robert Kane, MD, Harvey Solomon, MD, Erin Phillips, Laurie Ferrer, Cardinal Glennon Children's Hospital, St. Louis University, St. Louis, MO; Thomas Heffron, MD, Jill DePaolo, Todd Pillen, Laurel Davis, RN, Children's Healthcare of Atlanta, Atlanta, GA; John Bucuvalas, MD, Fred Ryckman, MD, Andre Hawkins, Gajra Arya, Children's Hospital, Cincinnati, Cincinnati, OH; Michael R. Narkewicz, MD, Ronald J. Sokol, MD, Frederick Karrer, MD, Cara Mark, MD, Kathy Orban-Eller, RN, MS, Children's Hospital Denver, University of Colorado School of Medicine, Denver, CO; Abhi Humar, MD, Brenda Durand, RN, Leslie Studenski, MPH, University of Minnesota, Minneapolis, MN; Elizabeth Rand, MD, Kathleen Anderer, CRNP, Children's Hospital of Philadelphia, Philadelphia, PA; George Mazariegos, MD, Nydia Chien, RN, MSN, Lynn Seward, RN, Children's Hospital of Pittsburgh, Thomas E Starzl Transplantation Institute, Pittsburgh, PA; Paul Atkison, MD, PhD, Children's Hospital Western Ontario, London, Ontario, Canada; Jay Roden, MD, Naveen Mittal, MD, Lisa Cutright, RN, CPNP, Children's Medical Center of Dallas, Dallas, TX; Grzegorz Telega, MD, Stacee Lerret, CPNP, Children's Hospital of Wisconsin, Milwaukee, WI; Estelle M. Alonso, MD, Joan Lokar, APN/NP, Susan Kelly, RN, BSN, Katie Neighbors, MPH, Children's Memorial Medical Center, Transplant Center of Excellence, Chicago, IL; Walter S. Andrews, MD, James Daniel, MD, Vicki Fioravanti, RN, CCTC, Angela Tendick, RN, BSN, Children's Mercy Hospital, Kansas City, MO; Anne S. Lindblad, PhD, Ravinder Anand, PhD, Changhong Song, PhD, Karen Martz, MS, Jeff Mitchell, Gladys Fraser, Nicole Hornbeak, Nirali Patel, Jianghang He, The EMMES Corporation, Rockville, MD; Annie Fecteau, MD, Vicky Lee Ng, MD, Maria De Angelis, NP, Andreanne Benidir, Hospital for Sick Children, Toronto, Toronto, Ontario, Canada; Kathleen Schwarz, MD, Paul Colombani, MD, May Kay Alford, MSN, CPNP, Michelle Felix, MSN, CPNP, Robert Jurao, Johns Hopkins Hospital, Baltimore, MD; James Eason, MD, John Eshun, MD, Sandra L. Powell, RN, MSN, Le Bonheur Children's Medical Center, Memphis, TN; Deborah Freese, MD, Jody Weckwerth, RN, Jean Greseth, RN, Lori Young, RN, Mayo Medical School, Rochester, MN; Robert Fisher, MD, Michael Akyeampong, RN, Medical College of Virginia, Richmond, VA; Samuel So, MD, William Berquist, MD, Marcia Castillo, RN, Annalie Bula, Stanford University Med Center, Palo Alto, CA; Sukru Emre, MD, Benjamin Shneider, MD, Nanda Kerkar, MD, Salvador Cuellar, CRC, Mount Sinai Medical Center, New York, NY; Saul Karpen, MD, PhD, Jaymee Mayo, RN, BSN, John Goss, MD, Douglas Fishman, MD, Val McLin, MD, Beth Carter, MD, Christine O'Mahony, MD, Thomas Aloia, MD, Donna Garner, PNP, Texas Children's Hospital, Houston, TX; Susan Gilmour, MD, MSc, FRCPC, Bernadette Dodd, RN, BScN, Norman Kneteman, MD, MSc, FRCSC, University of Alberta, Edmonton; Sue McDiarmid, MD, Susan Fiest, RN, BSN, CCRC, UCLA Medical Center, Los Angeles, CA; Steven Lobritto, MD, Lesley Smith, MD, Patricia Harren, NP, Kristin Maseda, New York Presbyterian Hospital, New York, NY; Fernando Alvarez, MD, Steven Martin, MD, Carol Viau, RN, Sainte-Justine Hospital, Montreal, Quebec, Canada; Ross Shepherd, MD, Jeffrey Lowell, MD, Michelle Nadler, RN, St. Louis Children's Hospital, St. Louis, MO; Joel Lavine, MD, Ajai Khanna, MD, Rosemarie Clawson, RN, University of California, San Diego Med Center, San Diego, CA; James Lopez, MD, PhD, John Magee, MD, Vicky Shieck, RN, BSN, CCTN, University of Michigan, Ann Arbor, MI; Jeffrey Fair, Steven Lichtman, Ken Andreoni, Joanne Prinzhorn, Valorie Buchholz, University of North Carolina, Chapel Hill, Chapel Hill, NC; Simon Horslen, Melissa Young, University of Washington, Seattle, WA; Glenn Halff, MD, Stacey Wallace, University of Texas, HSC San Antonia, San Antonio, TX; Joseph Tector, MD, Joel Lim, MD, Jean Molleston, MD, Jean Pearson, Indiana University Medical Center, Indianápolis, IN; Humberto Soriano, MD, Kathleen Falkenstein, PhD, PNP, St Christopher's, Philadelphia, PA Andreas Tzakis, MD, Tomoaki Kato, MD, Debbie Weppler, RN, MSN, Lisa Cooper, RN, Monica Gonzalez, RN, BSN, Alma Santiago, University of Miami/Jackson Memorial, Miami, FL; Munci Kalayoglu, MD, Anthony D'Alessandro, MD, Nissa Erickson (CI), Robert Judo, Stuart Knechtle, MD, Elizabeth Spaith, RN, University of Wisconsin, Madison, WI; Regino Gonzalez-Peralta, MD, Marcia Hodik, RN, Stacia McCracken, ARNP, University of Florida – Shands, Gainesville, FL, Philip Rosenthal, MD, Danusia Filipowski, MD, University of California, San Francisco, San Francisco, CA; Linda Book, MD, Molly O'Gorman, MD, Cynthia K. Kawai, MS, APRN, Lacey Bruschke, BSN, RN, Jennifer Kraus, RN, Primary Children's Medical Center, Salt Lake City, UT; Alan Langnas, DO, Dean Antonson, MD, Jean Botha, MB, Bch, Wendy Grant, MD, Debra Sudan, MD, Debb Andersen, RN, Beverly Fleckten, BS, CCRC, Kris Seipel, BS, University of Nebraska Medical Center, Omaha, NB; J. Michael Millis, MD, Patricia Boone, RN, MSN, University of Chicago, Chicago, IL; Dev M. Desai, MD, Sherri Javis, RN, Duke University Medical Center, Durham, NC; Peter Abt, MD, Tomi Shisler, FNP, Cindy Mack, RN, University of Rochester Medical Center, Rochester, NY.