The field of intestinal transplantation (ITx) has improved dramatically over the past two decades. Initially, success was measured by single case experiences (1–3). Thereafter, the field has witnessed steady growth and improvement. The reasons underpinning this success are multifactorial including advancements in surgical technique, immunosuppression, antimicrobial drug therapy, and experience gained. However, several factors have limited progress. First, clinically successful ITx came relatively late compared with other solid-organ transplants, mostly as a result of graft immunogenicity and inadequate immunosuppression. Second, the decade of the 1990s was focused on improving short-term patient survival with improvements from less than 50% to approximately 80%. Finally, standardization of care and analysis were limited by the relative infrequency of ITx and ITx centers compared with other-solid organ transplants.
To combat these limitations, an Intestinal Transplant Registry (ITR) was established (4–6). Updated biannually, ITR represents the largest number of ITx recipients in existence with the 2009 iteration reporting 2188 ITx from 73 centers (7). Analysis showed that survival was affected by graft type, pretransplant location, era, and center volume. Despite its advantages in case numbers, these data are limited by the multicenter nature of registry data and the limited number of data points collected.
Several large transplant centers have spearheaded advancements in the field (8–15). Single-center data, although limited by patient numbers, is advantaged by uniform care protocols and more in depth analysis. However, analyses are restricted to operative and postoperative factors. There are no studies examining only pretransplant factors that have a potential major impact on outcome.
The purpose of this study was to evaluate the impact of preoperative and intraoperative variables on survival. Our data demonstrate several factors that strongly influence outcome and are potentially subject to modification before ITx.
During the study interval, 88 patients received 106 ITx. The characteristics of the patient population are shown in Table 1. The mean wait time from listing to transplant was 152±202 days.
The graft types were liver-intestine grafts (LIG) 62%, isolated intestinal grafts (IIG) 24%, multivisceral grafts (MVG) 10%, and modified MVG (mMVG) 4%. The ischemia times in hours were 6.7±2.1 (cold ischemia time [CIT]), 0.76±0.31 (warm ischemia time [WIT]), and 7.5±2.0 (total ischemia time). The estimated blood loss was 91±98 mL/kg. Twenty-four percent of patients underwent recipient splenectomy. A total of 12.3% had a preoperative panel reactive antibody (PRA) positive, 78.3% were negative, and 9.4% were unknown. Donor-specific antibody (DSA) data were as follows: 7.6% positive, 49.1% negative, and 43.3% unknown. T-cell and B-cell crossmatch (XM) were 6.6% and 3.7% positive, 78.3% and 75.5% negative, and 15.1% and 20.8% unknown, respectively. The donor spleen management was as follows: 66.7% removed before transplant and 33.3% removed after reperfusion. Immunotherapy included induction with no agent (11%), interleukin-2 receptor antagonist (IL2RA; 60%), or rabbit antithymocyte globulin (rATG)/OKT3 (29%). The 1-, 3-, and 5-year survival rates were 80%, 70%, and 65% (patient) and 74%, 68%, and 64% (graft), respectively (Fig. 1). The leading causes of patient death were infection (58%) and neurological (16%). The leading causes of graft loss were rejection (32.5%) and infection (10%). A total of 45% of the grafts were functional at the time of patient death.
Univariate predictors (Table 2) of improved graft survival were liver inclusive grafts, PRA less than 20%, absence of DSA, negative T-cell XM, evaluation weight less than 20 kg, pediatric age group, WIT less than 60 min, IL2RA induction, female gender, calculated glomerular filtration rate more than 90 mL/min/1.73 m2, the more recent era, and donor spleen reperfusion. On multivariate analysis, the two independent predictors were absence of DSA and the use of liver inclusive grafts. As can be seen by the Figure 2(A), 1-year graft survival was more than 90% when neither of these predictors were present and 0% when both were present.
Univariate predictors (Table 2) of improved patient survival were liver inclusive grafts, PRA less than 20%, the absence of DSA, negative T-cell XM, WIT less than 60 min, the more recent era, conservation of the recipient spleen, IL2RA induction, and CIT less than 10 hr. On multivariate analysis, the two independent predictors of outcome were absence of DSA and conservation of the recipient spleen. As can be seen by the Figure 2(B), 1-year graft survival was more than 90% when neither of these predictors were present and 0% when both were present.
This study represents one of the largest single-center reports on ITx spanning nearly two decades of experience. During this interval, outcomes have improved. Era separated at 2000 was an important determinant of outcome as has been reported elsewhere (7, 8). This has reasonably been attributed to the accrual of experience, refinements in surgical technique, and advancements in immunotherapy. In addition, long-term results are improving as shown by the 5-year survival reported herein. This stands in stark contrast to the results from the ITR and United Network for Organ Sharing (UNOS)/Organ Procurement and Transplant Network (OPTN) where long-term overall survival is less than 40% (6, 16).The reasons for this difference are less clear but may be related to the multicenter nature of registry data and center-specific experience.
The purpose herein was to examine specific variables that are predictive of patient and graft survival rather than simply reporting a larger experience with this uncommon transplant procedure. Indeed, advancements should be made by dissecting transplant experiences to determine practices and techniques that result in improved outcomes and those that do not. Until, recently, no single center had a significant number of patients to afford meaningful statistical analysis. Analyses looked to multicenter registries such as the ITR and UNOS/OPTN data. However, variable reporting and limited data points tended to diminish the value of the larger patient numbers that existed in these databases.
No published study has solely addressed pretransplant factors and outcomes and instead combined postoperative data points with preoperative and intraoperative variables. Abu-Elmagd et al. (8) reported their experience with 500 ITx with part of the analysis focused on pre-ITx variables. Outcomes were improved by shorter lengths of total parenteral nutrition (TPN) therapy (IIG only), era, age, graft type (LIG), and immunosuppression. The era was the most significant factor determining overall graft failure. Kato et al. (10) examined the outcomes of 143 pediatric ITx cases in 2006. Significant predictors of poor patient survival were graft type (IIG and LIG), induction therapy (not receiving IL2RA), pretransplant location (hospital), and age (<1 year). Goulet et al. (14) analyzed the survival of 52 pediatric patients undergoing ITx. Multiple past abdominal surgeries, fungal sepsis, steroid resistant rejection, pretransplant location (hospital), retransplantation, inferior vena cava (IVC) thrombosis, and the diagnosis of intestinal pseudoobstruction were associated with worse outcome. Using the latest 5-year data (7), the ITR found that predictors of survival were era, center volume, pretransplant status (home), and liver inclusive grafts. Lao et al. (17) analyzed 852 patients younger than 21 years who underwent ITx from the UNOS/OPTN database. Improved survival was found only based on the cause of intestinal failure.
Given these limitations, we focused our study only on peritransplant factors. Our univariate results revealed several interesting parameters affecting outcome including renal function, gender, IL2RA induction therapy, PRA, and operative technical factors. The finding that pretransplant renal function affects outcome reinforces our earlier report where post-ITx renal dysfunction adversely impacted outcome (18). Gender has not consistently been reported as an outcome predictor although Pittsburgh recently noted that females have a reduced risk of graft loss to rejection (8). We and others (10) have reported improved results with the use of IL2RA induction therapy although there is potentially an era bias in our results as most of these cases occurred after 2000. Finally, technical factors such as ischemia times are shown to have an effect on outcome.
We found three factors that independently predicted survival—DSA (patient and graft), liver-inclusive graft (graft), and recipient splenectomy (patient). Better outcomes were seen when a recipient did not have DSA or a splenectomy and when a liver inclusive graft was transplanted. Equally important, each of these factors is potentially subject to manipulation before transplant.
The role of tissue typing, preformed antibodies, and XM is ill defined in ITx. This may be attributable to the fact that many ITx programs are liver-based programs, and tissue typing has not traditionally played a significant role in outcomes after liver transplantation. Newer data are however emerging that requires rethinking this position. Bond et al. (19) first reported on the effect of a positive XM on outcome. They found that the incidence, number, and severity of rejection were all higher in the XM-positive group, particularly IIG recipients. More recently on larger analysis, they report that a positive XM did not affect survival (8). In our data, a PRA more than 20%, DSA, and a positive XM match all adversely affected outcome at the univariate level. Only DSA affected both patient and graft survival on multivariate analysis. As a result, we have implemented strategies to reduce PRA and avoid human leukocyte antigen antibodies to which the recipient is presensitized.
Recipient splenectomy is a common practice both during and after ITx. Reasons for this procedure include technical (as part of native foregut resection), abdominal compartment space limitations, hypersplenism, and immunologic. Overwhelming postsplenectomy sepsis is of course a major concern particularly in immunosuppressed children. The data to date on recipient splenectomy is limited. Pittsburgh has reported a higher incidence of posttransplant lymphoproliferative disorder in splenectomized recipients (20). More recently, recipient splenectomy was associated with a lower risk of graft failure caused by rejection (8). In our study, recipient splenectomy was associated with a worse patient survival presumably because of sepsis.
An interesting corollary is the role of the donor spleen. Spurred by a clinical observation of long-term graft survival and the absence of rejection in a recipient of an mMVG with temporary retention of the donor spleen in 2000, we adopted a protocol wherein the implantation of all pancreas-inclusive allografts would entail the revascularization of the spleen for at least 1 hr. The concept was that although the bone marrow augmentation studies from Pittsburgh did not show an effect on rejection or graft survival (8), splenic reperfusion may augment these endpoints. Our data show that on univariate analysis, reperfusion of the donor spleen is associated with an improved graft survival, but this was not an independent predictor of survival. The University of Miami has reported 60 ITx recipients who received a donor spleen transplant (spleen was not removed after reperfusion) (21) and compared them with 81 who did not. There was no difference in survival, graft versus host disease, number of rejection, or infectious episodes. There was a significant difference in freedom from rejection and the authors concluded that inclusion of the spleen may provide long-term benefit and does not adversely impact short-term outcome. Our data would seem to support these conclusions.
Finally, the role of the liver in ITx outcomes is now becoming clearer. For years, the liver was purported to exert a protective effect on intestinal grafts without supporting clinical data. Pittsburgh first reported that patients receiving liver-inclusive ITx had a significantly lower cumulative risk of graft loss from rejection (22). Paris reported a significantly higher graft survival rate for LIG recipients when compared with IIG recipients (23). Using larger patient datasets, both Pittsburgh (8) and the ITR (7) have shown that inclusion of the liver is protective. Our data complement these studies on the protective effect of the liver.
Rejection was not included as an endpoint or as an outcome parameter in this study as it cannot be predicted or manipulated before transplantation. However, undoubtedly, it is an important factor in outcomes in this patient population as is attested by the fact that the most common cause of graft loss was rejection. Further, many of our pretransplant variables such as DSA and nonliver inclusive grafts imply that rejection is playing a major role here. In one of the most detailed publications on this subject, Selvaggi et al. (24) from the University of Miami examined rejection after primary ITx in 209 recipients. Univariate statistical analyses revealed that era of ITx, absence of Campath-1H induction therapy, failure to include the donor spleen, long CIT, and absence of any induction therapy all predicted rejection. On multivariate analysis, only transplant era was predictive of rejection. Interestingly, female recipients (more likely to be presensitized) and a positive XM did not influence rejection. Because the factors that influence rejection in this field are poorly understood, further investigations in this area are cleared needed.
Although this study has the advantages of being a single-center large data analysis, there are limitations that must be discussed. First, the fact is that this study spans approximately two decades of experience. Indeed, patient care and therapies have evolved during this time undoubtedly effecting analysis. Second, there is an era bias here as the majority of ITx have occurred after 2000, and the standard use of IL2RA occurred after that time. Further, although the patient numbers are large for an ITx series, a 100 plus graft analysis does carry some statistical limitations. One important limitation along these lines is the fact that a significant number of patients did not have DSA investigated before ITx. This gap in data results mostly from our pre-2000 experience and our evolution in tissue typing in this patient population.
Still our experience demonstrates that both short- and long-term outcomes after ITx have improved dramatically. Further, analysis focused solely on peritransplant factors shows that survival is adversely impacted by the presence of DSA, the use of a nonliver-inclusive allograft, and recipient splenectomy. Efforts to manipulate and improve these variables before transplantation should result in further improvements in outcome.
MATERIALS AND METHODS
A retrospective review using a prospectively maintained database of all ITx recipients transplanted between November 1991 and February 2010 was performed. Institutional Review Board approval of the database and analysis were undertaken before initiation of this study.
Over this nearly 20-year experience, management protocols have evolved. Herein are the general management strategies used:
- Candidates for ITx were assessed in a multidisciplinary manner and conformed to standardized criteria for ITx including the presence of irreversible intestinal failure with one or more parenteral nutrition-associated complications (25, 26). Patients were registered with the UNOS/OPTN.
- Donors were ABO identical or compatible. Target size matching was defined as a donor:recipient weight ratio less than 1. The donor operation is conducted using standard techniques as described (27).
- The immunogenetics have evolved with advancement in methodology. Retrospective B- and T-cell donor-recipient XMs were performed in all cases. Initially, cytotoxic XM techniques were used, which is now supplemented by flow cytometric XM techniques. Screening recipient PRA is routine. The initial PRA was complement-dependent lymphocytotoxicity antibody testing but is now Luminex (solid phase) bead testing. The antibodies identified are used to determine the presence of DSA. Desensitization protocols have been in place since 1999 using a combination of plasmapheresis, intravenous immunoglobulin, and rituximab (Rituxan, Genentech, South San Francisco, CA).
- The conduct of the recipient operations has been described elsewhere (28). Isolated intestinal grafts (IIG) include the donor jejunoileum and are transplanted orthotopically with arterial inflow from the infrarenal aorta or superior mesenteric artery α-actin and venous outflow through the superior mesenteric vein (SMV), splenic vein, or infrarenal IVC. Standard gastrointestinal reconstruction is undertaken including the use of ileostomies for graft monitoring. The liver-intestinal graft (LIG) consist of the liver and jejunoileum. Preservation of the donor pancreaticoduodenal complex has been routinely used since 2000 (29). The grafts are implanted orthotopically with preservation of the native foregut and arterial inflow from the supraceliac aorta, retention of the retrohepatic IVC, and venous outflow through the suprahepatic IVC. An end-to-side native portocaval shunt is created for drainage of the native foregut. Gastrointestinal continuity is restored as mentioned earlier. The multivisceral graft (MVG) include all or part of the liver, duodenum, pancreas, spleen, and jejunoileum. The stomach is not routinely included, and the native foregut is removed. Vascular anastomoses and gastrointestinal continuity is completed as mentioned earlier. The modified multivisceral graft are similar to the MVG except the liver is not included. It may be implanted orthotopically with removal of the native foregut or heterotopically as described for IIG. Since 2000, for all LIG, MVT, and mMVG, the donor spleen was reperfused for at least 60 min and then removed. Recipient splenectomy was performed on a case-by-case basis.
- Immunosuppression: nearly all patients have been managed with a tacrolimus (Prograf, Astellas USA, Deerfield, IL)-based regimen. All receive steroids that are weaned off in 1 to 3 years. Third agents are standard and include azathioprine (Imuran, Prometheus Laboratories, San Diego, CA; 1991-1995), cyclophosphamide (Cytoxan, Brystol Meyers Squib, New York, NY; 1996-1999), and mycophenolate mofetil (Cellcept, Genentech; 1999 to present). There are four eras of induction therapy: muromonab-cd3 (Orthoclone OKT3, Janssen-Cilag/Johnson & Johnson, New Brunswick, NJ; 1991-1995), no induction therapy (1995-1999), IL2RA (daclizumab/Zenapax, Roche Pharmaceuticals, Nutley, NJ or basiliximab/Simulect, Novartis Pharmaceuticals, East Hanover, NJ; 1999 to present), or antilymphocyte product (rATG/Thymoglobulin, Genzyme, Cambridge, MA; 2000 to present). Rejection is treated with steroid pulse/taper or antibody therapy (OKT3 or rATG). Positive XMs are treated with intravenous immunoglobulin, plasmapheresis, or rituximab.
Forty-four variables were analyzed. Preoperative variables were assessed at two time points—candidate evaluation (evaluation) and before ITx (preoperative). Variables were broadly classified into demographic (n=6), pretransplant characteristics (n=16), pretransplant laboratory data (n=12), and perioperative data (n=10) (Table 3). Adults and children were separated at age 18 years. The remnant bowel length was estimated from surgical measurements or as a calculated percentage of remaining gut using 400 cm as standard. Limited vascular access was defined is loss of more than or equal to three central venous access sites. Model for End-Stage Liver Disease/Pediatric End-Stage Liver Disease score is calculated from the UNOS calculator (30) with the actual score resulting from direct calculation while the adjusted score reflecting the match run score (exception points included). The clinical degree of liver disease was based on physical, pathologic, and laboratory assessment and divided into ESLD (cirrhosis with evidence of portal hypertension), EARLY (evidence of reversible liver disease in absence of significant portal hypertension), and NONE (no evidence of clinically significant liver disease). The glomerular filtration rate was calculated using Cockcroft-Gault formula for adults (31) and the Schwartz formula for children (32). Era was divided at the year 2000. WIT was calculated from removal of allograft from ice storage to vascular reperfusion, CIT as donor cross-clamp to removal of the allograft from ice storage, and total ischemia time was the sum of WIT+CIT. Liver inclusive grafts were LIG and MVG.
The primary endpoints of the study were overall patient and death-censored graft survival. The calculated means and medians of continuous variables were used as the points of differentiation. Some continuous variables were also organized into clinically significant thresholds. Patient and graft survival were calculated using Kaplan-Meier methods. Graft survival was death censored. Univariate analysis was conducted using log-rank test for categorical variables and Cox proportional hazard model for continuous variables. Variables that were significant at the P less than or equal to 0.20 univariate level were included in multivariate analysis. Model reduction was performed using the backward elimination variable selection method. P value less than or equal to 0.05 was considered significant. Statistical analysis was performed using JMP, version 8.0 (SAS Corp, Cary, NC). The median follow-up time was 35 months (range 0.1-180 months).
The authors thank Ms. Gladys Ruiz for her assistance with the preparation of this article.
1. Starzl TE, Rowe MI, Todo S, et al. Transplantation of multiple abdominal viscera. JAMA
1989; 261: 1449.
2. Grant D, Wall W, Mimeault R, et al. Successful small-bowel/liver transplantation. Lancet
1990; 335: 181.
3. Deltz E, Schroeder P, Gundlach M, et al. Successful clinical small-bowel transplantation. Transplant Proc
1990; 22: 2501.
4. Grant D. Current results of intestinal transplantation. The International Intestinal Transplant Registry. Lancet
1996; 347: 1801.
5. Grant D. Intestinal transplantation: 1997 report of the international registry. Intestinal Transplant Registry. Transplantation
1999; 67: 1061.
6. Grant D, Abu-Elmagd K, Reyes J, et al; Intestine Transplant Registry. 2003 report of the intestine transplant registry: A new era has dawned. Ann Surg
2005; 241: 607.
7. Intestinal Transplant Registry 2009. http://intestinaltransplantassociation.com/
. Accessed August 20, 2010.
8. Abu-Elmagd KM, Costa G, Bond GJ, et al. Five hundred intestinal and multivisceral transplantations at a single center: Major advances with new challenges. Ann Surg
2009; 250: 567.
9. Reyes J, Mazariegos GV, Bond GM, et al. Pediatric intestinal transplantation: Historical notes, principles and controversies. Pediatr Transplant
2002; 6: 193.
10. Kato T, Tzakis AG, Selvaggi G, et al. Intestinal and multivisceral transplantation in children. Ann Surg
2006; 243: 756.
11. Tzakis AG, Kato T, Levi DM, et al. 100 multivisceral transplants at a single center. Ann Surg
2005; 242: 480.
12. Langnas A, Chinnakotla S, Sudan D, et al. Intestinal transplantation at the University of Nebraska Medical Center: 1990 to 2001. Transplant Proc
2002; 34: 958.
13. Fishbein TM, Kaufman SS, Florman SS, et al. Isolated intestinal transplantation: Proof of clinical efficacy. Transplantation
2003; 76: 636.
14. Goulet O, Sauvat F, Ruemmele F, et al. Results of the Paris program: Ten years of pediatric intestinal transplantation. Transplant Proc
2005; 37: 1667.
15. Farmer DG, McDiarmid SV, Yersiz H, et al. Outcome after intestinal transplantation: Results from one center's 9-year experience; discussion 1031-2. Arch Surg
2001; 136: 1027.
16. Mazariegos GV, Steffick DE, Horslen S, et al. Intestine transplantation in the United States, 1999-2008. Am J Transplant
2010; 10(4 pt 2): 1020.
17. Lao OB, Healey PJ, Perkins JD, et al. Outcomes
in children after intestinal transplant. Pediatrics
2010; 125: e550.
18. Watson MJ, Venick RS, Kaldas F, et al. Renal function impacts outcomes
after intestinal transplantation. Transplantation
2008; 86: 117.
19. Bond G, Reyes J, Mazariegos G, et al. The impact of positive T-cell lymphocytotoxic crossmatch on intestinal allograft rejection and survival. Transplant Proc
2000; 32: 1197.
20. Abu-Elmagd K, Reyes J, Todo S, et al. Clinical intestinal transplantation: New perspectives and immunologic considerations. J Am Coll Surg
1998; 186: 512.
21. Kato T, Tzakis AG, Selvaggi G, et al. Transplantation of the spleen: Effect of splenic allograft in human multivisceral transplantation. Ann Surg
2007; 246: 436.
22. Abu-Elmagd K, Reyes J, Bond G, et al. Clinical intestinal transplantation: A decade of experience at a single center. Ann Surg
2001; 234: 404.
23. Goulet O, Lacaille F, Colomb V, et al. Intestinal transplantation in children: Paris experience. Transplant Proc
2002; 34: 1887.
24. Selvaggi G, Gaynor JJ, Moon J, et al. Analysis of acute cellular rejection episodes in recipients of primary intestinal transplantation: A single center, 11-year experience. Am J Transplant
2007; 7: 1249.
25. Kaufman SS, Atkinson JB, Bianchi A, et al; American Society of Transplantation. Indications for pediatric intestinal transplantation: A position paper of the American Society of Transplantation. Pediatr Transplant
2001; 5: 80.
26. American Gastroenterological Association. American Gastroenterological Association medical position statement: Short bowel syndrome and intestinal transplantation. Gastroenterology
2003; 124: 1105.
27. Yersiz H, Renz JF, Hisatake GM, et al. Multivisceral and isolated intestinal procurement techniques. Liver Transpl
2003; 9: 881.
28. Saggi BH, Farmer DG, Yersiz H, et al. Surgical advances in liver and bowel transplantation. Anesthesiol Clin North America
2004; 22: 713.
29. Sudan DL, Iyer KR, Deroover A, et al. A new technique for combined liver/small intestinal transplantation. Transplantation
2001; 72: 1846.
. Accessed August 20, 2010.