Errata: Erratum Volume 13, Issue 9, 2509, Article first published online: 4 September 2013
Outcomes of intestinal transplants (ITx; n = 977) for pediatric patients are examined using the United Network for Organ Sharing data from 1987 to 2009. Recipients were divided into four age groups: (1) <2 years of age (n = 569), (2) 2–6 years (n = 219), (3) 6–12 years (n = 121) and (4) 12–18 years (n = 68). Of 977 ITx, 287 (29.4%) were isolated ITx and 690 (70.6%) were liver and ITx (L-ITx). Patient survival for isolated ITx at 1, 3 and 5 years, 85.3%, 71.3% and 65.0%, respectively, was significantly better than L-ITx, 68.4%, 57.0% and 51.4%, respectively, (p = 0.0001); this was true for all age groups, except for patients <2 years of age. The difference in graft survival between isolated ITx and L-ITx was significant at 1 and 3 years (Wilcoxon test, p = 0.0012). After attrition analysis of graft survival of patients who survived past first year, 3 and 5 years, graft survival for L-ITx patient was significantly better than those for isolated ITx. Isolated ITx should be considered early before the onset of liver disease in children >2 with intestinal failure but is not advantageous in patients <2 years.
In children, intestinal failure (IF) is defined as the reduction of functional gut mass below the minimum needed for digestion and absorption of nutrients and fluids required for growth (1). Multidisciplinary medical and surgical treatment is the key for successful intestinal adaptation. Medical strategies include prevention of line infection, control of bacterial overgrowth, optimization of parenteral nutrition (PN), early enteral feeding and prevention of IF-associated liver disease (IFALD; Ref. 2). Despite appropriate multidisciplinary treatment of IF, 20–40% of children with IF remain PN-dependent and those who developed PN complications, including line sepsis and liver dysfunction, are considered for an intestinal transplant (ITx; Refs. 3,4). A major challenge is to determine the correct time for referral of these patients for ITx. If referred early, that is, before the onset of liver disease these children will require only isolated ITx, whereas later in the disease process they will require a combined liver and intestine transplant (L-ITx).
Kato et al., in a study of pediatric patients, reported that successful isolated ITx before the onset of liver failure resulted in better early outcomes, decreased the length of hospitalization and allowed more rapid conversion to enteral feeding (5). In a single-center retrospective study of pediatric recipients, Jugie et al., suggested that concomitant liver transplantation might improve small bowel graft and patient survival, although no statistical significance was noted for both graft and patient survival, when comparing children who received L-ITx to isolated ITx (6). They also showed that the rate of acute rejection was higher in isolated ITx when compared to composite grafts containing the liver (6). Berg et al., analyzing 465 patients, reported better patient survival for isolated ITx recipients as compared to combined L-ITx recipients, however the data was not separated for adult and pediatric patients (7). It may be postulated that adult and pediatric outcomes are different due to a variety of reasons including a difference in underlying etiology and comorbidities that are common in adults. We here analyzed data from the Organ Procurement Transplant Network/United Network of Organ Sharing (OPTN/UNOS) database specifically for pediatric patients who underwent primary ITx and compared the outcomes of isolated ITx to that of L-ITx to examine the beneficial role of either over the other. In addition the physiology of the children changes with age. An individual having gone through puberty at 17 years of age is biologically an adult, and therefore to combine outcomes for this age group with children <2 years old can yield erroneous inferences. We therefore specifically divided the population in this analysis. Most of the above mentioned series have focused on 1-, 3- and 5-year survival of these patients, however as known, the first year is the most critical for the recipient of multivisceral transplants and hence in this analysis we specifically tried to evaluate the long term survival of patients who survive past 1 year.
Materials and Methods
Using the OPTN/UNOS dataset, we evaluated data for all pediatric patients (<18 years old at the time of transplant) who underwent an ITx from October 1987 through August 2009. As with all retrospective studies using the UNOS datasets it is important to explain that the data is not absolutely complete; nevertheless, it remains the most comprehensive information in the United States. We followed the rules of the institutional review board of the University of Arizona for analysis of de-identified data. We defined an isolated ITx as a recipient of a small intestinal graft but not the liver and a L-ITx recipient as a patient who received a liver graft along with the small intestine. Some patients from both groups received additional abdominal organs, such as the stomach, kidney and/or pancreas; however, we did not further categorize the groups based on these organs.
We divided our study period arbitrarily into two eras: Era 1, October 1987 through December 2000, and Era 2, January 2001 through August 2009. The survival rates of ITx appeared to improve up until 2000, leading to approval by the Centers for Medicare & Medicaid Services (CMS) for coverage of ITx in 2000. Therefore, our two eras allow us to compare prior practices with current practice. In view of variations in indications, size and different biological and physiological characteristics of different ages, we divided the recipients arbitrarily into the following groups: (1) <2 years of age, (2) 2–6 years, (3) 6–12 years and (4) 12–18 years.
We calculated descriptive statistics for patient and donor characteristics. For categorical variables, we used the chi-square test or the Fisher's exact test; the Wilcoxon two-sample test or the Kruskal–Wallis tests were used to analyze continuous variables. The Kaplan–Meier method was used to compute patient survival and graft survival rates; for comparison, we used the Wilcoxon (short-term) and log-rank (long-term) test. When we refer to graft survival in patients who had more than an isolated ITx, we are referring to the survival of the intestinal graft. The Cox proportional hazards regression model was used for multivariate analysis and results are reported in terms of relative risk. When evaluating laboratory tests, we used the typical maximum value of our laboratory to define abnormal values. For all analyses, a p-value ≤0.05 was considered statistically significant. All analyses were performed with SAS 9.2 (SAS Institute Inc., Cary, NC, USA).
Of 977 primary ITx performed during the study period, 287 (29.4%) underwent an isolated ITx and 690(70.6%) an L-ITx. Five hundred and 69(58.2%) patients were in group 1, 219(22.4%) in group 2, 121(12.4%) in group 3 and 68(6.7%) in group 4. Among these patients, 86(15.1%), 88(40.2%), 71(58.7%) and 42(61.8%) had isolated ITx in groups 1, 2, 3 and 4, respectively, whereas 483(84.9%), 131(59.8%), 50(41.3%) and 26(38.2%) had L-ITx (percentages are calculated for each type of transplant within a group), respectively. The majority of L-ITx was performed in patients <2 years of age, corresponding to 70% of all L-ITx cases. In contrast, the distribution of isolated ITx was more uniform throughout the different age groups. The detailed demography and indications of transplantation are described in Table 1. Other organs transplanted with an isolated intestine were kidneys in four patients and pancreas in 19 patients. In the L-ITx group, 338 patients received a pancreas as part of the composite graft and another 31 patients received kidneys.
Table 1. Patient demographics and characteristics
FBP = functional bowel problems; NEC = necrotizing enterocolitis; SGS = short gut syndrome.
By age groups (years)
Indication for transplantation
Microvillous inclusion disease
Neuronal intestinal dysplasia
Protein losing enteropathy
The overall patient survival for primary isolated ITx recipients was 85.3%, 71.3% and 65.0% at 1, 3 and 5 years, respectively, which was significantly better than the survival of L-ITx recipients, 68.4%, 57.0% and 51.4%, respectively (p = 0.0001). The analysis of age-adjusted patient survival for patients undergoing isolated ITx and L-ITx are as shown in Table 2 and their statistical significance is described in Table 3. Survival of patients <2 years was significantly inferior to the older age groups for patients undergoing isolated ITx but not for L-ITx. The results were consistent even when we grouped all the patients between 2 and 18 years of age together and compared them with <2-year-old children. We did this because approximately 50% of children were in <2-year category. When patient survival was analyzed in the two eras, this statistically significant difference was observed in era 1 (p = 0.0057) and era 2 (p = 0.0002) (Table 4). The analysis of age-adjusted graft survival for patients undergoing isolated ITx and L-ITx are shown in Table 2 and 3 and in the different eras it is shown in Table 4.
Table 2. Age-adjusted patient and graft survival for isolated intestinal transplant versus liver-intestinal transplant before and after attrition
Age groups (years)
*For 1 and 3 years.
ITx = intestinal transplants; L-ITx = liver and intestine transplant.
Age groups (years)
Patient survival postattrition
Graft survival postattrition
Table 3. Patient and graft survival for isolated intestinal transplant versus liver-intestinal transplant comparing different age categories
ITx = intestinal transplants; L-ITx = liver and intestinal transplant.
Table 4. Age-adjusted patient and graft survival for isolated intestinal transplant versus liver and intestinal transplant separated in era1(1987–2000) and in era 2(2001–2009)
ITx = intestinal transplants; L-ITx = liver and intestine transplant.
ERA 1 (1987–2000)
44.5 (p = 0.0057)
41.3 (p = 0.06)
ERA 2 (2001–2009)
53.8 (p = 0.002)
48.6 (p = 0.010)
To examine the effect of attrition on patient survival we calculated patient survival for patients who had survived the first year. The patient survival for primary isolated ITx was 83.7% at 3 years and 75.8% at 5 years and for L-ITx it was 82.7% and 75.9%, respectively. This difference was statistically not significant (p = 0.6). When we analyzed patient survival in the different age groups no difference was observed as compared to that seen without attrition (Table 2).
Cause of death
The leading cause of death was infection for patients undergoing isolated ITx (27.4%). Other causes were: multiorgan failure (15%), graft failure (6.8%), respiratory failure (5.4%), cardiovascular (4.1%), postoperative bleeding (4.1%), posttransplant lymphoproliferative disease (PTLD; 4.1%) and cerebrovascular disorders (2.7%). For L-ITx, causes of death were infection (36.9%), multiorgan failure (14.7%), graft failure (6.3%), respiratory failure (4.5%), cardiovascular (4.9%), postoperative bleeding (4.9%), PTLD (2.4%) and cerebrovascular disorders (2.1%). For any particular cause of death, no statistical difference was found between L-ITx and isolated ITx.
The graft survival for primary isolated ITx recipients was 78.7%, 59.1% and 47.7%, at 1, 3 and 5 years, respectively and for L-ITx recipients, 66%, 54% and 47.2%, respectively. The difference in the graft survival rate between isolated ITx and L-ITx was significant at 1 and 3 years posttransplant but not at 5 years (Wilcoxon test, p = 0.0012; Figure 1A). During era 1, the difference in the graft survival rate between isolated ITx and L-ITx recipients was not significant at 1 and 3 years posttransplant (Wilcoxon test, p = 0.06; Table 4). In era 2, the difference in the graft survival rate between isolated ITx and L-ITx recipients was statistically significant at 1 and 3 years posttransplant (Wilcoxon test, p = 0.010; Table 4). The analysis of age-adjusted survival for patients undergoing isolated ITx and L-ITx are shown in Tables 2 and 3 and in the different eras it is shown in Table 4.
To examine the effect of attrition on graft survival we calculated graft survival for patients who had survived the first year. The graft survival was 75.6% and 60.6% at 3 and 5 years for patients undergoing isolated ITx and 81.7% and 72.6% for L-ITx recipients, respectively. Graft survival for L-ITx was significantly better than primary isolated ITx, (p = 0.03) contrary to the overall graft survival without attrition as described earlier (Figure 1B). In the age group analysis, graft survival was significantly better for L-ITx than isolated ITx for patients <2 years old, but not for the other 3 groups. Unlike in the overall analysis of graft survival, where graft survival was poor for babies <2 years old undergoing isolated ITX compared to the older age groups, after attrition this difference was not observed.
Cause of graft failure
The causes of graft failure for patients undergoing isolated ITx were reported as acute rejection (24.7%), chronic rejection (13.3%), infection (8.9%), technical problems (6.2%), PTLD (5.4%) and other causes (4.5%). In 37% it was reported that the patients “died with a functioning graft” (DWFG). For patients who underwent L-ITx the reported causes were, acute rejection (8.6%), infection (7.4%), technical problems (3.4%), chronic rejection (2.3%), PTLD (1.6%), other causes (6.6%) and DWFG (reported in 70.1%).
Cox regression analysis was performed for patient survival of the following parameters; transplant era, medical condition at the time of the transplant, age, gender, serum albumin, total bilirubin, Epstein Barr virus status before transplant, short bowel syndrome versus functional bowel disease and other organs transplanted. For patients undergoing isolated ITx aged <2 years (HR 3.0 [CI 1.6–5.6]), intensive care unit (ICU) admission at the time of transplant (HR 1.8 [CI 0.7–4.4]) was predictive of a worse outcome for patient survival. For patients undergoing L-ITx, ICU admission at the time of transplant (HR 1.9 [CI 1.4–2.6]) and serum albumin >4 (HR 1.9 [CI 1.2–3.2]) had a significant negative impact on patient survival.
Our findings indicate that the patient survival for primary isolated ITx was better than L-ITx. However this result did not hold true for children <2 years old. The poor graft survival rates of isolated ITx patients indicate that those <2 years of age may eventually require retransplantation. Furthermore there is a high probability of developing progressive liver disease as a consequence of long-term PN. These individuals would need an L-ITx graft as their second transplant. The results of L-ITx as a retransplant in children, especially <2 years old have been shown to be extremely poor (8). Previous studies have also shown retransplantation as negative predictor of patient outcome (9). Hence, even though the overall patient survival of isolated ITx is better than L-ITx, it can be argued that children <2 years old should receive their first transplant as L-ITx.
In this analysis, the graft survival was not significantly better at 5 years for isolated ITx as compared to L-ITx for the overall cohort of pediatric patients, even though the patient survival was better. The longer patient survival can be explained by the fact that if an isolated ITx graft fails, a graft enterectomy can be done and the patient can be placed back on PN. Of note, in the analysis by Sauvat et al., the graft survival for the combined graft was significantly higher than for isolated grafts and their L-ITx graft survival was much higher at 3 years (9). In the recent large series from Pittsburgh, the 5-year actuarial patient survival was 81% and graft survival was 76% for combined L-ITx recipients with rabbit anti-human thymocyte globulin induction, however a benefit was not seen in isolated ITx (10). All these reports are suggestive of protective effect of a liver graft. Also, from the heart transplant literature, it has been shown that the liver has a protective effect against rejection of the heart if transplanted simultaneously (11). Hence to test our hypothesis further when we removed all the patients who died within first year and then analyzed the graft survival, it was significantly better at 3 and 5 years for L-ITx graft compared to isolated ITx. This was true for the overall cohort of patients as well as children <2 years old in isolation (Table 2). Of note, these attrition results were very different than those observed in the adult population (12).
Fishbein in a review noted that ITx outcome was better in patients admitted from home, in younger patients, in patients undergoing their first transplant and in patients treated with antibody induction therapy (13). Other factors reported in the literature for poor outcome were intestinal pseudo-obstruction, multiple pretransplant surgeries, previous inferior vena cava thrombosis, exfoliative rejection, fungal sepsis, retransplants and abdominal wall complications (8). In our regression analysis, we found that patient admitted to ICU before transplantation had the worst outcome for both isolated ITx and L-ITx. We found that increased albumin >4g/dL was associated with a poor outcome in L-ITx. This may seem contradictory, but it can be inferred that these patients must have been the sickest requiring large amounts of exogenous albumin.
In conclusion, patient survival was better for isolated ITx than for L-ITx, except for patients <2 years of age. The better outcome for patients who received an isolated ITx improved further as the patient's age increased. This fact was not observed for patients who had L-ITx. Hence, it may be prudent to strongly consider children <2 years of age for an L-ITx as the preferred transplant whereas older children should be referred earlier for isolated ITx especially in view of improving results in era 2. This does however raise a question that should children <2 years old requiring isolated ITx wait until the liver disease progresses so that they can get a combined transplant or should they receive a preemptive liver transplant? This is a major debate in this area and is beyond the scope of this database and this manuscript. It also involves significant societal and financial implications. At the same time, because ICU admission at the time of transplant was the worst predictor for outcome after ITx, it is imperative to recognize which pediatric patients are too sick to transplant to increase the availability of organs to those recipients who will have better chance of good outcome.
The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation. This work was supported in part by the Health Resources and Services Administration contract 231–00-0115. Data reported here has been taken from the OPTN/UNOS as of August 25, 2009. The content is the responsibility of the authors alone and does not necessarily reflect the views and policies of the Department of Health and Human Services, nor does the mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government.