In contrast to other transplanted organs, the early detection of pancreas allograft rejection is elusive. Despite advances in immunosuppression, the incidence of acute rejection after solitary pancreas transplantation remains high (40% to 60%). Serologic markers of rejection may be unreliable, and histopathologic examination is the gold standard for the diagnosis of pancreas rejection.
Dr. Papadimitriou and colleagues of the University of Maryland, Baltimore, reported on a revised histologic grading scheme for chronic pancreas allograft rejection. This group had previously proposed a 5-point histologic grading scheme for chronic rejection based principally on the degree of pancreatic fibrosis. However, in the original scheme, grades II and III were difficult to reproduce, often inconsistent with true pathologic progression, and more representative of random sample variation. For this reason, they redefined the morphologic criteria and proposed an updated grading scheme for chronic rejection of the pancreas. In order to evaluate the new grading scheme, 2 pathologists performed a blinded analysis of 130 pancreas allograft biopsies from 36 randomly selected patients and graded them using a revised scheme as follows:
Grade 0 – No fibrosis
Grade I – Mild septal fibrosis in less than 30% of core surface
Grade II – Moderate septal fibrosis in 30% to 60% of core surface
Grade III – Severe septal fibrosis in greater than 60% of core surface
The authors noted that acinar atrophy was often proportional to septal fibrosis. In addition, in samples that contained vessels, transplant arteriopathy was likewise proportional to fibrosis. The updated grades of chronic rejection were then correlated with graft outcome. The average times from transplant to biopsy according to grade were significantly different (P = .0001), as shown below.
Grade 0 – 4 months
Grade I – 5.7 months
Grade II – 10.8 months
Grade III – 18.2 months
The average times from biopsy to graft failure according to grade were likewise significantly different (P = .001), as shown below.
Grade 0 – 41 months
Grade I – 22 months
Grade II – 10.8 months
Grade III – 1.2 months
The reproducibility between pathologists for each grade was quite high. The proposed updated scheme correlates well with graft outcome, is easily reproducible, and predicts, to a large extent, the remaining time of graft function.
Experience with surveillance pancreas allograft biopsy monitoring in solitary pancreas transplant recipients was reported by Dr. Larson and colleagues of the Mayo Clinic in Rochester, Minnesota. The study population consisted of 29 solitary pancreas transplant recipients, including 17 pancreas alone and 12 sequential PAK transplants. The overall 1-year pancreas graft survival rate in this group was 89%. The immunosuppressive protocol included tacrolimus (TAC), mycophenolate mofetil (MMF), and prednisone. All patients received antibody induction, and the authors divided their experience into 3 groups based on antibody induction: (1) daclizumab (n = 6), (2) OKT3 (n = 10), and (3) thymoglobulin (n = 13). Their biopsy protocol included a surveillance pancreas biopsy at 7 days after the last dose of antibody induction drug followed by surveillance biopsies at 2-3 months and at 12 months after transplantation. In addition, biopsies were performed for clinical indications such as an unexplained rise in the serum amylase or lipase or a reduction in the urine amylase for patients with bladder drainage. A total of 89 biopsies were performed in this patient population, and 7 complications arose: 2 hematomas, 2 patients with hematuria, 1 arteriovenous fistula, 1 patient with abdominal pain, and 1 oversedation. The majority of the biopsies were performed as surveillance biopsies.
The incidence of occult rejection detected on surveillance biopsies performed in the first 3 months after transplantation ranged from 15% to 22%, depending on the timing of the biopsy. For clinically indicated biopsies, the incidence of acute rejection was 77%. In the 6 patients receiving daclizumab induction, 3 (50%) experienced acute rejection; all 3 were diagnosed on surveillance biopsies. In the 10 patients receiving OKT3 induction, 5 (50%) experienced acute rejection; 4 cases were detected on surveillance biopsies. In the 13 patients receiving thymoglobulin induction, only 1 acute rejection occurred, and this was diagnosed from a clinically indicated biopsy. Based on this experience, surveillance biopsies were considered useful following solitary pancreas transplantation, particularly in patients receiving either daclizumab or OKT3 induction therapy. In general, these biopsies can be performed safely and biopsy results do not necessarily correlate with biochemical abnormalities of pancreas inflammation or dysfunction.
Dr. Marsh and colleagues of the University of Washington, Seattle, analyzed their overall experience with 332 pancreas biopsies during an 8-year period. From 1991 through 1995, pancreas transplant recipients were managed with CsA-based therapy; from 1995 through 1998, TAC-based immunosuppression was used. Their biopsy protocol included surveillance biopsies on days 10, 21, and 40, and 1 year after transplantation. In addition, biopsies were performed for clinical indications. Complications included 15 cases of bleeding (4.5%) and 10 cases of hematuria (3%).
The authors attempted to perform simultaneous kidney and pancreas biopsies whenever a biopsy was being performed. Three-year kidney and pancreas graft survival rates were 88%. Fifty percent of the biopsies were normal, and there was concordance between kidney and pancreas biopsies. In approximately 20% of cases, however, the kidney biopsy showed rejection and the pancreas biopsy was negative; and in approximately 10% of cases, the pancreas biopsy showed rejection and the kidney biopsy was normal. Therefore, the discordance rate was about 30%. Not surprisingly, the incidence of occult rejection was higher in biopsies performed earlier vs later. In addition, the overall incidence of acute rejection was lower in patients receiving TAC vs CsA. Notably, the 1-year surveillance kidney biopsies showed fibrosis in 28% of cases. However, biopsy findings of either rejection or fibrosis were not necessarily associated with reduced 3- and 5-year graft survival rates. Surprisingly, unexpected or occult mild rejection in both organs was found on the 1-year surveillance biopsy in several cases. The current protocol is to perform surveillance biopsies at 1, 6, and 12 months after SPK transplantation. The authors contend that surveillance pancreas biopsies, even in the setting of TAC-based therapy, enhance their ability to detect rejection and may improve long-term graft survival.
Pancreas transplantation is the preferred therapy for most patients with diabetes mellitus and end-stage renal disease. Advances in surgical techniques and immunosuppressive regimens have led to improved outcomes, yet formidable challenges continue to confront the pancreas transplant community. The most notable of these challenges are the shortage of donor organs, chronic rejection despite progress in preventing and treating acute rejection, and financial barriers that too often lead to patient noncompliance with obligatory immunosuppressant regimens and, ultimately, graft failure.
Much of the transplant community continues to focus on graft and patient survival as a primary outcome measure. This will undoubtedly continue to be the case until the problem of allograft rejection is resolved by highly specific immunosuppressive or tolerance induction strategies.
Data from the IPTR were analyzed by Dr. Angelika Gruessner and colleagues of the University of Minnesota, Minneapolis, to compare outcomes in 3 recipient groups: (1) SPK transplants from different donors; (2) SPK transplants from the same donor; and (3) sequential PAK transplants from different donors. The IPTR, located at the University of Minnesota in Minneapolis, is a scientific database collected from institutions performing pancreas transplants worldwide. In cooperation with over 200 centers, the pretransplant and posttransplant courses of over 10,000 patients who have received pancreas transplants are followed.
Since median waiting times for solitary pancreas transplants are less than half those for SPK transplants, an increasing number of PAK transplants are being performed, particularly after living donor kidney transplantation. SPK transplantation from different donors combines the advantages of a live donor kidney with only 1 surgical procedure for the recipient. Presumably, this would result in decreased waiting time and better use of organs from the available donor pool. However, a potential disadvantage is logistical and organizational problems related to performing a simultaneous live donor kidney and cadaver donor pancreas transplant.
Between January 1995 and July 2000, 57 SPK transplants from different donors, 5006 SPK transplants from the same donor, and 602 PAK transplants were reported to the IPTR. Only primary transplants were included in the analysis. In a multivariate model, the authors studied patient and graft survival rates, technical failure rates, graft loss rates from rejection, waiting times, outcome according to duct management, and basic demographics. In the SPK transplants from different donors, 61% were performed preemptively as compared with 22% in SPK transplants from the same donor. The mean waiting times were 195 days for SPK transplants from different donors, 150 days for PAK transplants, and 360 days for SPK transplants from the same donor. Length of initial hospital stay was 12 days in both SPK groups as compared with 9 days in the PAK group.
Patient survival was 95% in the PAK and SPK from same donor transplant groups, as compared with 88% in SPK transplants from different donors.
Kidney graft survival rates were 88% for SPK transplants from different donors, 90% for SPK transplants from the same donor, and 95% for PAK transplants. The death-censored kidney graft survival rates were 93%, 97%, and 99% for SPK transplants from different donors, SPK transplants from the same donor, and PAK transplants, respectively. The 1-year pancreas graft survival rates were 82% for SPK transplants from different donors, 83% for SPK transplants from the same donor, and 76% for PAK transplants (P < .001). However, at 2 years, the pancreas graft survival rate in SPK transplants from different donors approached the graft survival rate for PAK transplants, and was reduced compared with the 2-year pancreas graft survival rate in SPK transplants from the same donor.
The technical failure rate due to thrombosis was 5% in the 2 SPK groups as compared with 8% in the PAK transplant group. The 2-year rate of immunologic graft loss was lowest in the SPK transplant group from the same donor (2.6%). By contrast, the 2-year immunologic graft loss rate in the PAK group was 9%, and in the SPK transplant group from different donors, 10%. The incidence of rejection was 35% in SPK transplants from the same donor, 40% in PAK transplant recipients, and 50% in SPK transplants from different donors.
Based on this analysis, patient and kidney graft survival rates are comparable among groups. SPK transplantation from different donors significantly decreases waiting times. However, the best pancreas graft survival rates are still obtained for SPK transplants from the same donor. The rate of pancreas graft loss due to rejection is higher both in the PAK and SPK from different donor groups. However, graft loss from technical failure is lower for SPK from different donors compared with PAK. In the SPK transplants from different donors, the authors noted an increased rate of late acute rejection compared with SPK transplants from the same donor. Whether these late acute rejection episodes are abrogated by portal venous drainage is unknown at this time. SPK transplantation from different donors appears to be another viable option to increase the donor pool for Type 1 diabetic patients with end-stage renal disease who have the advantage of a voluntary live donor.
Analysis of United Network for Organ Sharing (UNOS) Scientific Registry: Acute Rejection and Long-Term Graft Survival
An analysis of data from the UNOS Scientific Registry was performed by Dr. Reddy and associates of the University of Kentucky, Lexington, to determine the impact of acute rejection episodes on long-term graft survival in SPK transplant recipients. All SPK and cadaver kidney-alone transplants reported to the UNOS database during 1988-1997 who had a diagnosis of Type 1 diabetes comprised the study population. Only patients whose kidney and pancreas grafts had survived for at least 1 year posttransplantation were included in the analysis. If the patient died with a functioning graft, the observation was censored. Other potential risk factors that can have an impact on the long-term graft survival were analyzed by a Cox Proportional Hazards model. Five-year graft survival rates were adjusted for risk factors, and half-lives were calculated.
A total of 4251 SPK and 8609 cadaver donor kidney transplant-alone recipients with Type 1 diabetes were identified. The SPK transplant group was characterized by younger donor recipient age, a lower level of sensitization, and reduced cold ischemia time. In the SPK transplant group, the overall incidence of acute rejection was 55% (36% kidney rejection alone, 3% pancreas rejection alone, and 16% both kidney and pancreas rejection). The incidence of acute rejection in kidney alone transplant recipients was 35%. In the SPK group, the 5-year kidney graft survival rates were 91% with no rejection, 88% with kidney rejection only, 94% with pancreas rejection only, and 86% with both kidney and pancreas rejection. The corresponding 5-year pancreas graft survival rates were 85% with no rejection, 84% with kidney rejection only, 83% with pancreas rejection only, and 78% with both kidney and pancreas rejection. There was a 42% increase in the relative risk of kidney graft failure in SPK recipients who had at least 1 rejection episode in the first-year posttransplantation. By comparison, kidney alone recipients who experienced at least 1 rejection episode in the first year after transplant had a 61% increase in graft failure. The difference in relative risk between the 2 groups was not significant. In the SPK group, the kidney graft half-life ranged from 34 years with no rejection to 20 years with both kidney and pancreas rejection. The pancreas graft half-lives ranged from 20 years with no rejection to 13 years with both kidney and pancreas rejection. Based on this analysis, the authors concluded that the occurrence of an early combined kidney and pancreas rejection episode had a negative impact on long-term kidney and pancreas graft survival. Acute kidney rejection did not affect long-term pancreas graft survival, while acute pancreas rejection did not affect the long-term kidney graft survival, suggesting that isolated kidney and pancreas rejection episodes may exist. Finally, the probability of graft failure associated with acute rejection in SPK transplant recipients was similar to the relative risk of graft failure in cadaver-donor kidney-alone transplant recipients who experienced at least 1 early rejection episode.
A multivariate analysis of a single-center experience with technically successful pancreas transplants performed between January 1994 through December 1999 was conducted by Dr. Humar and colleagues of the University of Minnesota, Minneapolis. A total of 542 pancreas transplants were analyzed: 212 SPK, 232 PAK, and 98 pancreas-alone transplants. The incidence of acute rejection was greatly influenced by the category of transplant. At 1-year posttransplantation, the overall incidence of acute rejection was 29% in SPK, 47% in PAK, and 61% in pancreas-alone recipients (P = .01). The incidence of pancreas graft loss due to acute rejection was 0.9% in SPK, 1.3% in PAK, and 3.1% in pancreas-alone transplant recipients. The incidence of pancreas graft loss due to chronic rejection was 2.8% in SPK, 9.9% in PAK, and 17.3% in pancreas-alone recipients (P < .05). Consequently, the 3-year pancreas graft survival rates were 92% in SPK, 85% in PAK, and 65% in pancreas-alone transplant recipients.
A multivariate analysis of the relative risk of pancreas graft failure was performed and acute rejection was identified as having the highest odds ratio of graft failure (5.4). Other risk factors for graft failure that achieved or approached significance included category of transplant, panel reactive antibody (PRA) titer above 20%, and recipient age above 45 years. A multivariate analysis for acute rejection was then performed in each of the 3 transplant categories. For SPK recipients, significant risk factors for acute rejection included non-TAC-based immunosuppression, mismatch at the DR loci, pancreas retransplantation, and non-MMF-based therapy. For PAK recipients, significant risk factors for acute rejection included a PRA titer above 20%, mismatching at the B locus, and non-TAC-based immunosuppression. For pancreas-alone recipients, the risk factors included non-TAC-based immunosuppression, recipient age below 45 years, and enteric drainage. Based on this analysis, acute rejection remains a significant problem after pancreas transplantation, especially in PAK and pancreas-alone transplant recipients.
Favorable experience with the use of “statins” in SPK transplant recipients was reported by Dr. Pesavento and colleagues of Ohio State University, Columbus. From 1987 to 2000, the authors studied 396 SPK transplant recipients after censoring for renal graft loss. A total of 99 of these patients were managed with statins, and 297 did not receive cholesterol-reduction therapy. The mean follow-up was 5 years and the mean time to the initiation of statin therapy was 3.3 years after transplantation. In patients receiving statins, the mean total cholesterol was 249 mg/dL prior to treatment as compared with 197 mg/dL in the patients not receiving statins. Correspondingly, the mean triglyceride levels at the time of initiation of statin therapy were 206 mg/dL vs 142 mg/dL in patients receiving vs not receiving statins. There was no evidence in either group of any significant differences in cardiovascular disease based on the pretransplant evaluation. With statin therapy, the mean total cholesterol level decreased to 200 mg/dL and triglycerides decreased to 160 mg/dL. Interestingly, in the 99 patients managed with statins, the overall patient survival was 90%, vs 65% in the remaining 297 patients not receiving statins (P < .05). Not surprisingly, cardiovascular events were the major cause of death in both groups. Even after censoring for early deaths, the patient survival advantage remained in the statin-treated group. The use of statins appears to have a profound impact on patient survival in SPK transplant recipients.
According to IPTR data, the majority of pancreas transplants are performed with systemic venous delivery of insulin, resulting in peripheral hyperinsulinemia. To improve physiology, a new technique of pancreas transplantation with portal venous drainage was developed. In animal models, portal venous delivery of antigen has been shown to induce tolerance.
UNOS registry data were also analyzed by Dr. Perez and colleagues of the University of California Davis Medical Center, Sacramento, to compare renal allograft rejection rates in SPK transplant recipients transplanted from 1994 to 2000 who underwent either portal or systemic venous drainage of the pancreas allograft. The analysis was restricted to recipients from pancreas transplant centers experienced in the technique of portal venous drainage. They identified 15 centers that had performed at least 7 pancreas transplants with portal venous drainage. A total of 909 SPK transplant recipients were analyzed from these centers, including 539 with systemic and 370 with portal venous drainage. The majority (67%) of portal venous drainage pancreas transplants were performed in the later era (1997-2000), while the majority of systemic venous drainage pancreas transplants (62%) were performed in the earlier era (1994-1996).
Both portal venous and systemic venous drainage groups were similar with regard to numerous donor and recipient demographic characteristics. However, the portal venous drainage group had a higher percentage of African-American recipients (16% vs 8% in the systemic drainage group), while the systemic venous drainage group had a higher percentage of patients with a PRA titer of 0 (62% vs 51% in the portal venous drainage group). Cold ischemia times were comparable.
The incidence of acute renal allograft rejection prior to hospital discharge was 16.4% in the portal venous vs 25% in systemic venous-drained SPK transplants (P < .001). The 6-month rates of acute renal allograft rejection were 35.6% with portal venous drainage vs 47.7% with systemic venous drainage (P < .001). The 12-month rates of rejection were also significantly lower in SPK recipients with portal venous vs systemic venous drainage. When rejection frequency was analyzed in patients transplanted in 1994-1996 vs 1997-2000, the difference between portal venous- and systemic venous-drained patients persisted in the later era but was more modest, possibly reflecting the availability of more potent immunosuppressive agents in recent years.
One-year kidney graft survival rates were not significantly different between portal (89%) and systemic (88%) venous drainage groups. These excellent kidney graft survival rates translated into a kidney half-life of 17 years with portal venous drainage and 15 years with systemic venous drainage of the pancreas (P = NS). Further analysis revealed a higher rate of acute renal allograft rejection in African-American recipients and in SPK patients with a PRA titer above 10%. In each of these higher-risk categories, patients with portal venous drainage had slightly lower rates of acute rejection compared with patients with systemic venous drainage. A multivariate analysis of risk factors for acute rejection revealed an odds ratio of 1.6 for systemic drainage, 2.3 for African-American recipients, and 2.8 for the earlier transplant era (1994-1996). Based on this analysis, portal venous drainage of the pancreas allograft results in less renal allograft rejection in SPK transplant recipients, possibly due to the induction of systemic hyporesponsiveness by portal venous delivery of antigen.
Dr. Philosophe and associates of the University of Maryland, Baltimore, performed a single-center retrospective study analyzing the rates of acute renal allograft rejection in SPK transplant patients receiving standardized immunosuppression according to technique of transplant. A total of 128 SPK transplants were performed between August 1995 and 2000: 62 had systemic venous drainage and 66 had portal venous drainage of the pancreas allograft. All patients received TAC-based therapy in combination with MMF, corticosteroids, and antibody induction. All episodes of renal allograft rejection were biopsy-proven.
Data were analyzed by Kaplan-Meier, univariate log rank, and multivariate Cox regression methodology. The 2 groups were quite similar with regard to numerous demographic characteristics. The rates of at least 1 or more pancreas rejection episodes in the first 24 months after transplantation were 10% with portal venous drainage vs 40% with systemic venous drainage (P < .05). The corresponding rates of acute renal allograft rejection were 28% with portal venous drainage vs 45% with systemic venous drainage of the pancreas (P < .05). The mean grade of rejection was similar between groups. Portal venous drainage had a calculated odds ratio of 0.50 with respect to the relative risk of acute renal rejection. The authors concluded that portal venous drainage of the pancreas in SPK transplant recipients lowers the incidence of acute rejection in the concomitant kidney and suggests a mechanism of immunologic protection based on portal-induced tolerance.