An intestinal transplant (ITx) has been increasingly accepted as the treatment of choice for patients with intestinal failure who have complications of total parenteral nutrition (1). Over the past decade, the number of such transplants performed in the United States has been gradually increasing (2–4). Although improved 5-year patient survival has been reported (5, 6), graft loss still remains a significant concern. It is unknown whether a high degree of graft loss in ITx is due to increased immunogenicity or due to poor definition and available testing to evaluate the graft and distinguish various causes for injury.
Historically, ITx retransplants have been associated with significant morbidity and mortality (7–9). More recent single-center data have shown better results (10). Magee et al. (3) reported liver, kidney, heart, and lung retransplant outcomes during the most recent decade in the United States; however, they did not address ITx retransplants.
In a pediatric series from Europe, an ITx retransplant was reported as a predictor of poor outcome (11). Unlike the International Transplant Registry, the Organ Procurement and Transplantation Network (OPTN) and the Scientific Registry of Transplant Recipients of the United Network for Organ Sharing (UNOS) represent a significant uniformity of data documentation, have a larger data set, and can potentially be used to implement policies in the United States. There is a limitation in that there is some missing data and a few fields will have higher reliability than others, but it cannot be denied that it remains a significant source of information about all these patients and used periodically for survival evaluation.
Of 1822 ITx in 1664 patients during our study period, 149 patients received retransplants (nine third transplants, all children). Of 149, 72 (48.3%) were in adults (8 of whom underwent primary transplants as children), and 77 (51.7%) were in children.
Of the 72 adult retransplants, 41 (57%) underwent an isolated ITx and 31 (33%) underwent a liver ITx (L-ITx). Of 41 isolated ITx retransplants, 39 (95.1%) had a previous isolated ITx and only 2 (4.9%) had a previous L-ITx. Of 31 L-ITx retransplants, 22 (71%) had a previous isolated ITx and 9 (29%) had a previous L-ITx.
Of the 77 pediatric retransplants, 28 (36%) underwent an isolated ITx and 49 (64%) underwent an L-ITx. Of 28 isolated ITx retransplants, 27 (96.4%) had a previous isolated ITx while 1 (3.6%) had a previous L-ITx. Of 49 L-ITx retransplants, 22 (44.9%) had a previous isolated ITx and 27 (55.1%) had a previous L-ITx.
By era, of the adult retransplants, 5 underwent an isolated ITx and 7 a L-ITx in era 1 and 36 underwent an isolated ITx and 24 a L-ITx in era 2. Of the pediatric retransplants, 10 underwent an isolated ITx and 15 a L-ITx in era 1 and 18 underwent an isolated ITx and 34 a L-ITx in era 2. Of the nine third transplants, one had an isolated ITx (11.2%) in era 2 and eight (88.8%) of them had a L-ITx (two in era 1 and six in era 2). Retransplant recipient characteristics are summarized in Table 1.
In the 41 adult isolated ITx retransplants, patient survival (same as graft survival) was 80.1% at 1 year posttransplant, 47.4% at 3 years, and 28.5% at 5 years; in the 31 adult L-ITx retransplants, survival was 63.1%, 56.1%, and 46.8%. The differences were not statistically significant by type of retransplanted graft.
For primary transplants, patient survival in adult isolated ITx was 84% at 1 year, 66.7% at 3 years, and 54.2% at 5 years; in adult L-ITx, survival was 67%, 53.3%, and 46%. The difference in patient survival after a primary transplant versus retransplant was statistically significant in adult isolated ITx (P=0.005) (Fig. 1a) but not in adult L-ITx. We specifically evaluated those 9 of 31 L-ITx recipients who had previous L-ITx and L-ITx as retransplant, 1, 3, and 5 year survival was 88%, 88%, and 77%. When compared with primary L-ITx, the difference was not statistically significant (P=0.07).
By era, patient survival in adult isolated ITx retransplants in era 1 (January 2001 through August 2009, n=5) was 60% at 1 year, 40% at 3 years, and 40% at 5 years; in era 2 (October 1987 through December 2000, n=36), it was 84.1%, 47.5%, and 16% (P=0.9). The patient survival in adult L-ITx in era 1 (n=7) was 28.6%, 28.6%, and 14.3% but increased to 73.6%, 64.4%, and 64.4% in era 2 (n=24) (P=0.01) (Fig. 2).
In the 28 pediatric isolated ITx retransplants, patient survival was 80.7% at 1 year, 74% at 3 years, and 57.5% at 5 years; in the 49 pediatric L-ITx, it was 42.1%, 42.1%, and 42.1% (P=0.03, log-rank test; 0.02, Wilcoxon test).
For primary transplants, patient survival in pediatric isolated ITx was 84.5% at 1 year, 70.8% at 3 years, and 64% at 5 years; in pediatric L-ITx, it was 67.6%, 56.1%, and 51.4%. The difference in patient survival in pediatric isolated ITx was not statistically significant between primary and retransplants (Fig. 1b). For pediatric L-ITx, it was significantly better after a primary transplant versus retransplant (P=0.0034). When we specifically looked at those 27 of 49 patients who had both first and second transplants as a L-ITx, 1, 3, and 5 year survival was 92%, 67%, and 50%. This difference was not statistically significant when compared with the primary L-ITx (P=0.5).
The difference in patient survival in pediatric recipients after a primary transplant versus a retransplant was statistically significant in era 2 for both types of transplant. In pediatric isolated ITx, in era 2, the patient survival was 83.8% at 1 year, 73.4% at 3 years, and 67.2% at 5 years after a primary transplant versus 76.5%, 76.5%, and 46.5% after a retransplant (P=0.0001). In pediatric L-ITx, patient survival was 71.6%, 60.1%, and 53.8% after a primary transplant versus 47.9%, 47.9%, and 47.9% after a retransplant (P=0.003). When the pediatric subgroups were analyzed based on their age, the only significant differences in outcome within subgroups were in the patient survival in L-ITx retransplants less than 2 years of age versus 10 to 18 years of age, as well as in those less than 2 years of age versus 18 years and older.The outcome in L-ITx retransplants less than 2 years of age was extremely poor: the patient survival at 1 year posttransplant was only 13%; the graft survival was the same (Fig. 3).
In the nine pediatric third transplants, the patient survival (same as the graft survival) after a third L-ITx (n=8) was 50% at 1 year, 18.8% at 3 years, and 18.8% at 5 years; the one third isolated ITx lived for only 3 months.
Of the 52 reported deaths among pediatric retransplants, the causes were bacterial sepsis (n=11), graft rejection (n=10), multiorgan dysfunction (n=9), fungal or other infections (n=8), and the rest were reported as others. Of the third transplants, four died; the reported cause was bacterial sepsis (n=2), fungal pneumonia (n=1), and fungal sepsis (n=1). The one-third isolated ITx died from severe acute rejection.
In the 41 adult isolated ITx retransplants, graft survival (same as the patient survival) was 80.1% at 1 year posttransplant, 47.4% at 3 years, and 28.5% at 5 years; in the 31 adult L-ITx retransplants, it was 63.1%, 56.1%, and 46.8%. The differences were not statistically significant by type of retransplant.
For primary transplants, graft survival in adult isolated ITx was 80.7% at 1 year, 56.7% at 3 years, and 42.8% at 5 years; in adult L-ITx, it was 64.1%, 51%, and 41.4%. The difference in graft survival after a primary transplant versus retransplant was statistically significant only for isolated ITx retransplant (P=0.03).
By era, graft survival in adult L-ITx retransplants significantly increased from 28.6% at 1 year, 28.6% at 3 years, and 14.3% at 5 years in era 1 to 73.6%, 64.4%, and 64.4% in era 2 (P=0.01). In adult isolated ITx retransplants, graft survival was 60%, 40%, and 40% in era 1 and 84.1%, 47.5%, and 16% in era 2 (P=0.9).
In the pediatric retransplants, graft survival at 1 and 3 years was significantly better after an isolated ITx than after an L-ITx (P=0.04, Wilcoxon test): the graft survival was 76.4%, 56.6%, and 44% in the 28 pediatric isolated ITx retransplants versus 39.3%, 39.3%, and 39.3% in the 49 pediatric L-ITx retransplants.
For primary transplants, graft survival in pediatric isolated ITx was 77.7% at 1 year, 58.6% at 3 years, and 47% at 5 years; in pediatric L-ITx, it was 65.1%, 53.1%, and 47.2%. The graft survival after a primary transplant was significantly better than after a retransplant only for pediatric L-ITx when all retransplanted L-ITx grafts were considered (P=0.0055). However, like patient survival, the difference between primary L-ITx grafts and those 27 of the 49 patients who had both L-ITx as first and second transplant was not significant; 1, 3, and 5 year survival was 92%, 67%, and 50% for these patients (P=0.4).
In era 1, graft survival in pediatric isolated ITx retransplants was 87.5% at 1 year, 50% at 3 years, and 50% at 5 years; in pediatric L-ITx retransplants, it was 26.7%, 26.7%, and 26.7%. In era 2, graft survival in pediatric isolated ITx retransplants was 70.1%, 62.3%, and 37.4%; in pediatric L-ITx retransplants, it was 42%, 42%, and 42%. We found no statistically significant differences by era.
However, the difference in era 2 in the graft survival after a primary transplant versus retransplant was statistically significant: in pediatric isolated ITx, the graft survival was 78.7% at 1 year, 62.8% at 3 years, and 48.5% at 5 years after a primary transplant versus 69.9%, 62.1%, and 38.1% after a retransplant (P=0.01); in pediatric L-ITx it was 68.9%, 57%, and 48.7% after a primary transplant versus 45.5%, 45.5%, and 45.5% after a retransplant (P=0.009). The major reported causes of graft failure (n=35) were acute rejection (n=17), chronic rejection (n=7), and infection (n=3).
In regression analysis, the only factor of significance was prior hospitalization (hazard ratio, 5.4; confidence interval [CI], 1.921–15.619) (Table 2). Other factors such as the panel reactive antibody (PRA) level of these patients that might have contributed were not consistently reported.
The retransplantation of any organ potentially deprives a patient of a primary transplant. Furthermore, solid-organ retransplant outcome has been reported to be questionable (3). Yet over the past decade, the number of retransplants has grown rapidly in all fields of solid-organ transplantation. In one of the most comprehensive review by Magee et al., they reported an increase in the absolute number of retransplants, most notably in kidney transplantation (3). In our study, intestinal retransplantation in adults increased almost five times in the past decade, whereas the overall number of primary ITx for adults had increased to 3.2 times (from 156 in era 1 to 506 in era 2). In the pediatric population, the number almost doubled for retransplant and had 2.4 times increase for primary transplants (from 283 in era 1 to 683 in era 2). The most notable increase was in adult isolated ITx retransplants, but these patients had the worst long-term patient and graft survival.
In kidney recipients, the mean time to retransplantation is approximately 5 years. In liver recipients it is usually less than 1 year because the liver is most likely to be lost to either primary nonfunction or hepatic artery thrombosis. In our isolated ITx recipients, the time to a retransplant ranged from 3 months to 1 year; this short time period reflects our finding that rejection is a driver for retransplantation of this organ.
We have shown that patient and graft survival for adult isolated ITx retransplants were inferior to primary transplants. There was, however, no such difference in pediatric isolated ITx. This result was similar to those from the single-center series reported by Mazariegos et al. (10). In this report, immunosuppression before the retransplant, the timing of the enterectomy, and a period of hyperalimentation after the retransplant were all considered important for outcome (10). In our analysis, for adults mainly, rejection was the major cause of graft failure. Humoral rejection is not well-defined in ITxs, and the poor results of isolated ITx retransplantation may be due to increased levels of the PRA and allosensitization. In a recent single-center series, Farmer et al. (12) showed improved outcome of ITx in patients with PRA less than 20%, absence of donor-specific antibody among other factors in univariate analysis. In kidney transplants, it is well known that graft nephrectomy is associated with an increase in PRA levels causing difficulties in obtaining a suitable subsequent match and resulting in poor outcome (13). The roles of the above mentioned factors, such as the timing of graft enterectomy and status of immunosuppression before retransplantation, cannot be addressed as they were not consistently reported in the database.
We found no statistically significant difference in patient and graft survival in adult L-ITx after a primary transplant versus retransplant. We therefore suggest that the caution expressed about retransplantation in adult L-ITx may be misplaced; on the contrary, L-ITx retransplantation should be considered, especially after a failed isolated ITx. Because isolated ITx graft loss is related to rejection, the absence of reduced graft survival in adult L-ITx retransplants indicates that sensitization and PRA levels may be less of a concern in the presence of immunogenic protection by the liver.
When evaluating results for L-ITx retransplants, we disregarded as well as regarded their first transplant type. It can be argued that the patient who receives an L-ITx after L-ITx is different to one receiving an L-ITx after a failed isolated ITx. It is a serious limitation of databases that the level of sickness of patients cannot be expressed. However, one who received L-ITx as a retransplant after failed isolated ITx will have progressive parenteral nutrition–associated liver disease because it can be safely assumed that patients with failing intestinal grafts could not have been enterally fed. And these patients will have progressive portal hypertension, whereas one with L-ITx after L-ITx may have absence of significant portal hypertension. The pediatric L-ITx retransplants were of interest in that the difference was significant when all patients were considered against those who had both L-ITx. We found that the outcome was worse in L-ITx retransplant recipients younger than 2 years. In that age group, we suppose that patient size and technical issues come into consideration.
In conclusion, patient and graft survival in isolated ITx recipients are less favorable after a retransplant compared with a primary transplant for adults. The patient and graft survival are also poor in pediatric L-ITx retransplants especially children younger than 2 years, as compared with a primary transplant. Long-term L-ITx retransplant results improved significantly in era 2 in adults. The most dismal long-term outcome was in adult isolated ITx retransplants and in infant L-ITx retransplants. Hospitalization immediately before the time of the retransplant is a predictor of poor outcome.
MATERIALS AND METHODS
Using the Standard Transplant Analysis and Research files of the UNOS, we evaluated data for all patients (pediatric patients, defined as <18 years of age) who underwent an ITx from October 1987 through August 2009. We further divided the pediatric age group into those younger than 2 years, 2 to less than 10 years, and 10 to less than 18 years. We did this because in these three age groups, there is significant variation in size and weight of recipient. As with all retrospective studies using OPTN and the Scientific Registry of Transplant Recipients data, it is important to emphasize that the data are not 100% complete; nevertheless, it remains the most comprehensive ITx information in the United States. We analyzed only those variables consistently reported in at least 90% of patients. We followed the rules of the University of Arizona Institutional Review Board for analysis of deidentified data. We defined an isolated ITx as a small intestinal graft without a liver graft and an L-ITx as a liver graft along with a small intestinal graft. Some recipients in either group received additional abdominal organs, stomach, kidney, and/or pancreas. These were not further categorized for analysis.
We divided our study period into two eras: era 1, October 1987 through December 2000, and era 2, January 2001 through August 2009. The patient and graft survival in ITx recipients seemed to improve until 2000, leading to Centers for Medicare and Medicaid Services approval for coverage as of 2001. Therefore, we considered January 2001 a suitable starting point for era 2, enabling us to assess the possible impact of practices since Centers for Medicare and Medicaid Services approval in 2001.
We calculated descriptive statistics, and used chi-square test, analysis of variance, or the Fisher exact test for categorical variables; for continuous variables, we used the Wilcoxon two-sample test or the Kruskal-Wallis test. The patient and graft survival rates were computed with the Kaplan-Meier method; to compare groups, the Wilcoxon (short-term) and log-rank (long-term) tests were used. For multivariate analysis, we used the Cox proportional hazards regression model. Results are reported as relative risk and corresponding P value. Abnormal laboratory parameters were taken from our laboratory's maximum values. We considered a P value less than or equal to 0.05 to be statistically significant using SAS 9.2 (SAS Institute Inc., Cary, NC) software.
Variables included were recipient ethnicity, human leukocyte antigen matching, a pretransplant albumin level less than 3 g/dL, a serum creatinine level more than 1.3 mg/dL, a total bilirubin level more than 1.3 mg/dL, hospitalization immediately before retransplantation, indication of transplant, time interval between two transplants, and donor age. We used a Cox regression model to evaluate significant findings for both patient and graft survival using these variables. Cutoff values for variables were decided based on our institutional cutoff values for abnormal number.
The authors thank Melissa M. Carton, B.A., for manuscript preparation and Mary Knatterud, Ph.D., for editing assistance.