Optimal timing for a pancreas transplant after a successful kidney transplant

BACKGROUND

For certain uremic, diabetic patients, a sequential transplant of a kidney (usually from a living donor) followed by a cadaver pancreas has become an attractive option. But how long to wait after the kidney transplant before proceeding with a pancreas transplant is unclear. We studied outcomes in recipients of a pancreas at varying times after a kidney to determine the optimal timing for the second transplant.

METHODS

We compared pancreas after kidney (PAK) transplants performed early (< or =4 months) and late (>4 months) after the kidney transplant to determine any significant differences in surgical complications or outcomes between the two groups.

RESULTS

Between January 1, 1994, and September 30, 1998, we performed 123 cadaver PAK transplants. Of these, 25 (20%) were early and 98 (80%) were late. Characteristics of the two recipient groups were similar. We found no significant differences in outcome between the two groups. The incidence of surgical complications (bleeding, leaks, thrombosis, infections) and of opportunistic infections (such as cytomegalovirus) did not significantly differ between the two groups. Graft and patient survival rates were also equivalent (P=NS). The incidence of acute rejection by 3 months posttransplant was 20% in both groups.

CONCLUSION

The timing of the pancreas transplant for PAK recipients does not seem to influence outcome. As long as an acceptable organ is available and the recipient is clinically stable, a PAK transplant can be performed relatively soon after the kidney transplant.

Decreased surgical risks of pancreas transplantation in the modern era

OBJECTIVE

To document the decreased incidence of surgical complications after pancreas transplantation in recent times.

SUMMARY BACKGROUND DATA

Compared with other abdominal transplants, pancreas transplants have historically had the highest incidence of surgical complications. However, over the past few years, the authors have noted a significant decrease in the incidence of surgical complications.

METHODS

The authors studied the incidence of early (<3 months after transplant) surgical complications (e.g., relaparotomy, thrombosis, infections, leaks) after 580 pancreas transplants performed during a 12-year period. Patients were analyzed and compared in two time groups: era 1 (June 1, 1985, to April 30, 1994, n = 367) and era 2 (May 1, 1994, to June 30, 1997, n = 213).

RESULTS

Overall, surgical complications were significantly reduced in era 2 compared with era 1. The relaparotomy rate decreased from 32.4% in era 1 to 18.8% in era 2. Significant risk factors for early relaparotomy were donor age older than 40 years and recipient obesity. Recipients with relaparotomy had significantly lower graft survival rates than those without relaparotomy, but patient survival rates were not significantly different. A major factor contributing to the lower relaparotomy rate in era 2 was a significant decrease in the incidence of graft thrombosis; the authors believe this lower incidence is due to the routine use of postoperative low-dose intravenous heparin and acetylsalicylic acid. The incidence of bleeding requiring relaparotomy did not differ between the two eras. Older donor age was the most significant risk factor for graft thrombosis. The incidence of intraabdominal infections significantly decreased between the two eras; this decrease may be due to improved prophylaxis regimens in the first postoperative week.

CONCLUSIONS

Although a retrospective study has its limits, the results of this study, the largest single-center experience to date, show a significant decrease in the surgical risk associated with pancreas transplants. Reasons for this decrease are identification of donor and recipient risk factors, better prophylaxis regimens, refinements in surgical technique, and improved immunosuppressive regimens. These improved results suggest that more widespread application of pancreas transplantation is warranted.

Steroid withdrawal in pancreas transplant recipients

BACKGROUND

Numerous studies of steroid withdrawal have been carried out in kidney and liver transplant recipients, but only a few in pancreas transplant recipients. Yet, pancreas transplant recipients could have significant long-term benefits from steroid withdrawal.

METHODS

We performed a retrospective analysis to determine the feasibility of steroid withdrawal in pancreas transplant recipients.

RESULTS

Of 360 recipients who underwent a pancreas transplant between January 1, 1994 and June 30, 1998, 14 attempted steroid withdrawal (12 simultaneous pancreas-kidney [SPK]; 2 pancreas transplant alone [PTA]). Reasons for steroid withdrawal were bone fractures (n = 3), psychiatric disorders (n = 2), severe acne (n = 1), recurrent infections (n = 4), and problems with hypercholesterolemia or hypertension (n = 4). All 14 were maintained on tacrolimus and mycophenolate mofetil (MMF) immunosuppression, except for 1 who was on tacrolimus and azathioprine (AZA). Of the 14 recipients, 11 had no episodes of acute rejection before steroid withdrawal. The remaining 3 had one or more acute rejection episodes. Of the 14 recipients, 10 (72%) currently remain off steroids (mean follow-up 18 months, range 5-51 months). However, 4 recipients have resumed steroids: 2 after an acute rejection episode (at 2 and 21 months post-withdrawal) and 2 because of leukopenia (WBC < 3000) and an inability to tolerate full-dose MMF. Steroid withdrawal was unsuccessful in both PTA recipients and in 2 of the 12 SPK recipients. All 14 recipients currently have a functioning pancreas graft. However, 1 of the SPK recipients, in whom steroid withdrawal failed, has developed chronic kidney rejection and is now back on hemodialysis awaiting a retransplant.

CONCLUSION

Steroid withdrawal is possible in up to 70% of pancreas transplant recipients. Further studies are necessary to define ideal candidates for steroid withdrawal. Based on the results of this analysis, we have launched a prospective, randomized trial of steroid withdrawal in pancreas transplant recipients.

Analysis of United States (US) and non-US pancreas transplants as reported to the International Pancreas Transplant Registry (IPTR) and to the United Network for Organ Sharing (UNOS)

As of November 1998, more than 11,000 pancreas transplants had been reported to the IPTR, including more than 8,800 US and more than 2,600 non-US cases. The 1994-98 cases (> 4,500) were analyzed, including more than 4,000 US and more than 500 non-US transplants. For all US 1994-98 SPK transplants (n = 3,409), one-year patient, pancreas and kidney graft survival rates were 94%, 90% and 83%, respectively; for all PAK cases (n = 375), one-year patient and graft survival rates were 95% and 71%; and for all PTA cases (n = 181), one-year patient and graft survival rates were 95% and 64%, respectively. Recipient age had only a small impact on outcome, with one-year patient survival rates for all recipients < 45 years of 95% (n = 3,215) versus 91% for those > or = 45 years old (n = 758) (p = 0.005). Pancreas graft survival rates at one year for those < 45 versus > or = 45 years old were 84% versus 78% in the SPK (p < 0.02), 70% versus 78% in the PAK (p = 0.13), and 62% versus 79% in the PTA (p = 0.23) categories, respectively). Nearly one-third of US pancreas transplants for 1994-98 were done by the ED drainage technique. For SPK transplants, the one-year pancreas graft survival rate was 83% for BD (n = 2,369) and 82% for ED (n = 912) (p < or = 0.09). For PAK and PTA transplants, pancreas graft survival rates were significantly higher with BD, 74% (n = 261) and 68% (n = 115), respectively, at one year. The drawback for BD was the need for conversion to ED in 7% of the cases at one year and 11% at 2 years. For TS transplants, the pancreas graft loss due to rejection was very low for SPK transplants, 2% at one year versus 9% for PAK and 15% for PTA cases. The various initial maintenance immunosuppressive regimens (Tac + MMF, Tac + Aza, CsA + MMF, CsA + Aza) resulted in only minor differences in pancreas graft survival rates in the SPK cases (80-86% at 1 year), but in PAK and PTA cases the Tac + MMF combination was associated with significantly higher pancreas graft survival rates. For BD PAK transplant recipients given Tac + MMF, the one-year pancreas graft survival rate was 83% (n = 100). For the corresponding BD PTA group it was 75% (n = 44). For non-US cases the outcomes were similar. For non-US SPK transplants (n = 586), one-year patient, kidney and pancreas graft survival rates were 93%, 85% and 81%, respectively. Cox multivariate analyses and logistical regression were done in each recipient category to assess the factors that influence pancreas graft loss. BD was associated with a significantly lower risk than ED in all categories. Increasing donor age was a risk factor in most categories. MMF was associated with a decreased risk for graft loss in the SPK category, and Tac in the PAK and PTA categories.

Pancreas transplants for United States (US) and non-US cases as reported to the International Pancreas Transplant Registry (IPTR) and to the United Network for Organ Sharing (UNOS)

As of November 1997 more than 9,800 pancreas transplants had been reported to the IPTR, including more than 7,400 US and more than 2,400 non-US cases. Cases reported since the inception of UNOS in 1987 (more than 6,800 in the US and more than 2,000 from outside the US) were analyzed, with emphasis on 1994-97 cases. In the US, for all 1994-97 SPK transplants (n = 2,585), one-year patient, pancreas and kidney graft survival rates were 94%, 82% and 90%, respectively; for all PAK cases (n = 230), patient and graft survival rates at one year were 95% and 71%; and for all PTA cases (n = 117), patient and graft survival rates at one year were 93% and 62%, respectively. The 1994-97 pancreas survival rates in all categories were significantly higher than in previous eras. Analysis of bladder versus enteric drainage (BD vs ED) was made only for 1994-97 SPK cases since there were too few solitary ED cases for a comparison. One-year pancreas survival rates (all causes of failure included) were 83% for BD (n = 1,995) versus 80% for ED (n = 456) cases (p = NS). However, nearly 20% of SPK BD grafts had been converted to ED by 2 years. The pancreas retransplant success rates have also significantly increased over time. For 1994-97 pancreas retransplants (n = 92), the one-year graft survival rate was 72%. The technical failure (TF) rate for pancreas transplants has declined over time. For 1994-96 BD cases, the TF rate was 8% in SPK (n = 1,995), 13% in PAK (n = 174) and 11% in PTA (n = 90) cases. For SPK ED cases (n = 456) the TF rate was 11% (p = 0.06 vs SPK BD). The most common cause of technical loss was graft thrombosis, 5.5% for SPK, 10.2% for PAK and 6.7% for PTA. The TF rates were significantly higher in grafts from older (> 45 yrs) donors and donors who died from cardiovascular disease. The immunological graft loss rate has also significantly declined in all categories. For 1994-97, technically successful (TS) cases (DWFG censored), the rejection loss rate at one year was 2% in the SPK (n = 2,234), 9% in the PAK (n = 154), and 16% in the PTA (n = 78) categories. Cox multivariate analyses were done in each recipient category to assess for factors that influence risks of pancreas graft loss. Increasing donor age was a risk factor in all categories. HLA mismatching was important only in the solitary (PAK and PTA) categories. FK506 use was associated with a positive effect in all categories, as was MMF in the SPK category.

Conversion from bladder to enteric drainage after pancreaticoduodenal transplantations

BACKGROUND

Bladder drainage is the most common technique for managing the exocrine secretions of pancreaticoduodenal grafts. However, bladder drainage can cause urinary, pancreatic, and metabolic complications that may require conversion to enteric drainage. With enteric drainage, urinary amylase levels cannot be monitored as a marker for rejection. After enteric conversion, rejection is the major cause of graft loss. Timing the conversion to reduce immunologic graft loss would greatly improve patient and graft survival rates. Our study was designed to assess the incidence of, indications for, and complications of converting from bladder to enteric drainage after pancreaticoduodenal transplantations.

METHODS

We retrospectively reviewed our experience with 80 recipients who underwent enteric conversion. We studied the recipient category, the interval from transplantation to conversion, the interval from the last rejection episode to conversion, the indications for conversion, the type of enteric drainage at conversion (loop versus Roux-en-Y), the results of the conversion, and postconversion complications.

RESULTS

The major indications for conversion were metabolic acidosis (n = 26, 33%), recurrent urinary tract infections (UTIs) (n = 16, 20%), reflux pancreatitis (n = 15, 19%), and hematuria (n = 12, 15%). For most recipients, their symptoms resolved after conversion (n = 76, 95%). The cumulative probability of undergoing conversion was 13% at 12 months, 21% at 36 months, and 25% at 60 months. Of the recipients with surgical complications after conversion (n = 12, 15%), one lost his graft as a result of pancreatitis. Overall, of the 80 recipients who underwent conversion, 12 (15%) lost their graft, most due to rejection (n = 8, 75%). Immunologic graft loss was highest for recipients of pancreas transplants alone who underwent conversion < or = 6 months after transplantation or < or = 1 year after their last rejection episode.

CONCLUSIONS

Enteric conversion is safe and therapeutic in recipients with complications related to the exocrine secretions of bladder-drained pancreas grafts. After conversion, rejection accounted for 75% of the grafts lost. However, waiting at least 1 year after the last rejection episode significantly reduced immunologic graft loss.

Mycophenolate mofetil in pancreas transplantation

BACKGROUND

Mycophenolate mofetil (MMF) has been shown to decrease the incidence of acute rejection episodes after kidney transplantation. The use of MMF along with tacrolimus for > or =1 year after pancreas transplantation has not been studied in a large single-center analysis.

METHODS

Between July 1, 1995 and June 30, 1997, both MMF and tacrolimus were given to 120 pancreas transplant recipients. By category, 61 underwent simultaneous pancreas-kidney transplantation (SPK); 44 underwent pancreas transplantation after previous kidney transplantation (PAK); and 15 underwent pancreas transplantation alone (PTA). By donor source, 86% of the grafts were from a cadaver, and 14% were from a living-related donor. Induction therapy was with MMF, tacrolimus, prednisone, and antithymocyte globulin (n=109) or OKT3 (n=2). Until oral intake was resumed, recipients initially received intravenous azathioprine. Side effects were as follows: gastrointestinal (GI) toxicity in 53% of recipients receiving combined MMF and tacrolimus therapy; bone marrow toxicity in 24% of recipients receiving MMF alone; nephrotoxicity in 18% and neurotoxicity in 11% of recipients receiving tacrolimus alone. We did a matched-pair analysis to compare outcome in MMF versus azathioprine recipients, using the database of the International Pancreas Transplant Registry. Matching criteria included transplantation category, transplantation number, recipient and donor age, duct management, HLA typing, and transplantation year.

RESULTS

One-year patient survival rates were 98% for SPK, 98% for PAK, and 100% for PTA (P=NS). For SPK recipients, 1-year pancreas graft survival rates were 86% with MMF versus 79% with azathioprine (P=NS); kidney graft survival rates were 96% with MMF versus 86% with azathioprine (P=NS). The incidence of first rejection episodes at 1 year was significantly lower for MMF recipients (15% with MMF versus 43% with azathioprine) (P = 0.0003). For recipients of solitary pancreas transplants (PTA and PAK), we found no difference in graft survival rates between MMF and azathioprine. The conversion rate from MMF to azathioprine at 1 year was 14% for SPK recipients, 26% for PAK, and 39% for PTA (P < 0.007). The most common reason for conversion was GI toxicity, in particular for nonuremic (PTA) or posturemic (PAK) recipients. The rates of posttransplant infection and lymphoproliferative disease were low for recipients on MMF and tacrolimus.

CONCLUSIONS

The combination of MMF and tacrolimus after pancreas transplantation is highly effective and safe. For SPK recipients, the incidence of acute reversible rejection episodes was significantly lower with MMF than with azathioprine. The conversion rate from MMF to azathioprine because of GI toxicity was lowest for SPK and highest for PTA recipients.

Pancreas transplantation: a review

Pancreas transplantation is now performed as a routine treatment for uremic diabetic recipients of kidney transplants either simultaneously with or after the kidney. Such patients are obligated to immunosuppression and with a successful pancreas transplant can achieve insulin independence as well as a dialysis-free state. Pancreas transplants alone are less commonly applied because of the need for immunosuppression, but the trade off to achieve an insulin-independent state may be worthwhile for individual patients, particularly those who are labile with hypoglycemic unawareness. This option should certainly be a part of the treatment armentation of the modern diabetologist. A positive effect on secondary complications will certainly occur with an early transplant, and even late can have an impact as has been shown for neuropathy. Whether the simpler procedure of islet transplantation will replace pancreas transplants remains to be seen. Of more than 200 islet allografts performed in the 1990s, less than 10% of the recipients have achieved insulin independence at 1 year. Clinical islet trials are ongoing but limited to patients who accept a low individual probability of success to assist in development, or to those in whom the surgical risks of a pancreas transplant is high. Islet transplantation has held promise for over 25 years, but candidates for endocrine replacement therapy must honestly be told the difference in success rates, which are currently much higher with the pancreas.

Donor-specific portal blood transfusion in intestinal transplantation: a prospective, preclinical large animal study

BACKGROUND

Unlike in kidney and heart transplantation, the role of pretransplant donor-specific blood transfusion (DST) has not been studied prospectively in a large animal model of bowel transplantation. We investigated the impact of portal versus systemic DST on overall survival, rejection, graft-versus-host disease (GVHD), and infection after total (small and large) bowel transplantation in pigs.

METHODS

Mixed lymphocyte culture-reactive, outbred pigs underwent total enterectomy and orthotopic total bowel transplantation with portal vein graft drainage. One unit of donor blood was transfused via the portal or systemic circulation (according to a randomization protocol) before graft implantation was begun. We studied six groups, all of which underwent at least a total bowel transplant: group 1 (n=5) comprised nonimmunosuppressed control pigs with portal DST; group 2 (n=6), nonimmunosuppressed control pigs with systemic DST; group 3 (n=5), cyclosporine (CsA)-treated pigs with portal DST; group 4 (n=5), CsA-treated pigs with systemic DST; group 5 (n=5), tacrolimus-treated pigs with portal DST; and group 6 (n=5), tacrolimus-treated pigs with systemic DST. All immunosuppressed pigs received prednisone (2 mg/kg/day) and either CsA (to maintain levels between 250 and 350 ng/ml) or tacrolimus (to maintain levels between 10 and 30 ng/ml). Stomal biopsies and autopsies were obtained to study the incidence of rejection, GVHD, and infection.

RESULTS

Portal DST and tacrolimus-based immunosuppression resulted in the highest survival rates. At 7, 14, and 28 days after transplantation, survival rates in group 5 were 100%, 100%, and 80%; in group 6, 100%, 60%, and 40%; and in group 3, 100%, 0%, and 0%, respectively. Only the combination of portal DST and tacrolimus prevented the occurrence of, and death from, rejection. Death from rejection at 7, 14, and 28 days in group 5 was 0%, 0%, and 0%; in group 6, 0%, 33%, and 67%; and in group 3, 0%, 100%, and 100%, respectively. Of note, if immunosuppression was used, the groups with portal (versus systemic) DST had a higher risk of death from infection but a lower risk of death from GVHD. Simultaneous immunologic events were noted more frequently in groups with systemic (versus portal) DST. Long-term survival was noted only in groups with tacrolimus-based immunosuppression and was more common for those with portal (versus systemic) DST.

CONCLUSIONS

Portal DST at the time of total bowel transplantation and posttransplant immunosuppression with tacrolimus prevent rejection and significantly increase graft survival. The combination of portal antigen presentation and tacrolimus needs to be studied in clinical bowel transplantation.

Pregnancy after pancreas transplantation in the cyclosporine era: report from the International Pancreas Transplant Registry

BACKGROUND

As of December 31, 1996, a total of 9012 pancreas transplants have been reported to the International Pancreas Transplant Registry (IPTR). Over 75% of these recipients were < or = 40 years of age at the time of their transplant. With continued improvement in short- and long-term outcome after transplant, an increasing number of female recipients have become pregnant. However, outcome after pregnancy in pancreas transplant recipients has not been studied in detail.

METHODS

To evaluate the risks of pregnancy after pancreas transplantation in the cyclosporine era, we surveyed the institutions reporting to the IPTR.

RESULTS

Nineteen cases of pregnancy in 17 female recipients of simultaneous pancreas-kidney transplants resulted in 19 live births. Metabolic control during pregnancy was good in all cases. Mean duration of gestation was 35.2-2.2 weeks. Mean birth weight was 2150+/-680 g. Two congenital malformations were reported (one bilateral cataract and one double aortic arch). One child developed type I diabetes at age 3 years. Only one pancreas graft and one kidney graft were lost (in two different recipients). The pancreas graft was lost after delivery (because of acute rejection). The kidney graft was lost 3 months after delivery (impaired function due to previous amphotericin B treatment). One case of a worsened secondary complication (retinopathy) was reported. One recipient died of myocardial infarction 7 years after transplant and 5 years after delivery, in spite of a normal pretransplant coronary angiogram and good pancreas function.

CONCLUSION

This study shows that simultaneous pancreas-kidney transplantation can restore fertility in uremic type I diabetic female recipients. Thus, both posttransplant contraception and fertility counseling are options for female recipients. However, our analysis demonstrates that, when discussing the possibility of pregnancy with female pancreas recipients, these potential risks must be considered: (1) graft loss, (2) progressive maternal morbidity and mortality even with good glycemic control, and (3) diabetes transmission to offspring.

Diagnosis of pancreas rejection: cystoscopic transduodenal versus percutaneous computed tomography scan-guided biopsy

BACKGROUND

The most common cause of graft failure after technically successful pancreas transplants is rejection. Laboratory parameters for detecting pancreas graft rejection are not consistently reliable and can lead to unnecessary antirejection treatment. Histopathologic evaluation is the gold standard in the differential diagnosis of pancreas graft dysfunction. Four biopsy techniques have been described: cystoscopic transduodenal (CB), percutaneous computed tomography scan-guided (PB), open, and laparoscopic biopsy.

METHODS

We studied the two most common techniques, CB and PB, in pancreas transplant recipients with presumed rejection. Group 1 comprised 103 attempts at CB in 82 recipients (53 men, 29 women) with bladder-drained (BD) pancreas transplants, at 1 to 80 (median, 14) months after transplant. Group 2 comprised 93 attempts at PB in 68 recipients (41 men, 27 women), at 0.5 to 64 (median, 14) months after transplant.

RESULTS

In group 1, of 103 attempts at CB, adequate tissue was obtained in 90 (87%): pancreas alone in 23 (22%), pancreas + duodenum in 35 (34%), duodenum alone in 32 (31%). Of the 58 pancreas biopsies, 23 (40%) showed acute rejection. Of the 67 duodenal biopsies, 16 (24%) showed acute rejection. Complications of CB included macrohematuria in 4 recipients (4%) and microhematuria in 32 (31%). We noted no biopsy-related pancreatitis. The mean cost of CB was $2561+/-246. In group 2, of 93 attempts at PB, adequate tissue (all pancreas) was obtained in 67 (72%); of these, 29 (43%) showed acute rejection. Of 23 inaccessible pancreases, 9 (39%) underwent CB; pancreatic tissue was obtained in four (45%), and results were consistent with rejection in all four. Complications of PB included biopsy-related pancreatitis (serum amylase > or = 25%) in five (7%) recipients, macrohematuria in one (1%), and abdominal hemorrhage in two (3%). The mean cost of PB was $1038+/-78. (1) CB and PB prevented unnecessary antirejection treatment in 44% of our recipients with successful biopsies; (2) CB had a higher success rate for obtaining tissue (including duodenal specimens) and a lower rate of major complications; (3) PB was easier and cheaper, did not require general anesthesia, and was performed as an outpatient procedure.

CONCLUSIONS

We conclude that PB should become the biopsy technique of choice in recipients with presumed pancreas graft rejection. If PB fails, recipients with bladder-drained pancreas transplants should undergo CB. If CB fails, or in recipients with enteric-drained or duct-injected pancreas transplants, a laparoscopic or open biopsy should be considered.

Surgical complications requiring early relaparotomy after pancreas transplantation: a multivariate risk factor and economic impact analysis of the cyclosporine era

OBJECTIVES

To study significant surgical complications requiring early (< or = 3 months posttransplant) relaparotomy (relap) after pancreas transplants, and to develop clinically relevant surgical and peritransplant decision-making guidelines for preventing and managing such complications.

SUMMARY BACKGROUND DATA

Pancreas grafts are still associated with the highest surgical complication rate of all routinely transplanted solid organs. However, the impact of surgical complications on morbidity, hospital costs, and graft and patient survival rates has not been analyzed in detail to date.

METHODS

We retrospectively studied surgical complications requiring relap in 441 consecutive cadaver, bladder-drained pancreas transplants (54% simultaneous pancreas and kidney [SPK]; 22% pancreas after kidney [PAK]; 24% pancreas transplant alone [PTA]; 37% retransplant). Outcome and hospital charges were analyzed separately for recipients with versus without reoperation.

RESULTS

The overall relap rate was 32% (SPK, 36%; PAK, 25%; PTA, 16%; p = 0.04). The most common causes were intraabdominal infection and graft pancreatitis (38%), pancreas graft thrombosis (27%), and anastomotic leak (15%). Perioperative relap mortality was 9%; transplant pancreatectomy was necessary in 57% of all recipients with one or more relaps. The pancreas graft was lost in 80% of recipients with versus 41% without relap (p < 0.0001). Patient survival rates were significantly lower (p < 0.05) for recipients with versus without relap. By multivariate analysis, significant risk factors for graft loss included older donor age (SPK, PAK), retransplant (PAK), relap for infection (SPK, PAK), and relap for leak or bleeding (PAK). For death, risk factors included older recipient age (SPK, PAK),retransplant (SPK, PAK), relap for thrombosis (PAK), relap for infection or leak (SPK), and relap for bleeding (PTA).

CONCLUSIONS

Posttransplant surgical complications requiring relap were frequent, resulted in a high rate of pancreas (SPK, PAK, PTA) and kidney (SPK, PAK) graft loss, and had a major economic impact (p = 0.0001). Complications were associated with substantial perioperative mortality and decreased patient survival rates. The focus must therefore shift from graft salvage to preservation of the recipient's life once a pancreas graft-related complication requiring relap occurs. Thus, the threshold for pancreatectomy should be low. In this context, acceptance of older donors and recipients must be reconsidered.

Solitary pancreas transplantation for nonuremic patients with labile insulin-dependent diabetes mellitus

BACKGROUND

Simultaneous pancreas-kidney transplantation has become a widely accepted treatment option for selected uremic patients with insulin-dependent diabetes mellitus (IDDM). Patient survival rates at 1 year exceed 90%, and rates of pancreas graft survival, 70%. However, solitary pancreas transplantation for nonuremic patients with IDDM has been controversial because of the less favorable outcome and the need for long-term immunosuppression with its associated morbidity and mortality.

METHODS

We studied the outcome of 225 solitary pancreas transplants during three immunosuppressive eras: the precyclosporine (CsA) era (n=83), the CsA era (n=118), and the tacrolimus era (n=24). Only patients with labile IDDM (e.g., hypoglycemic unawareness, insulin reactions, > or = 2 failed attempts at intensified insulin therapy for metabolic control) underwent solitary pancreas transplantation. Using univariate and multivariate analyses, we looked at patient and graft survival, the risk of surgical complications, and native kidney function during these three eras.

RESULTS

Pancreas graft survival improved significantly over time: 34% at 1 year after transplantation in the pre-CsA era, 52% in the CsA era, and 80% in the tacrolimus era (P=0.002). Pancreas graft loss due to rejection decreased from 50% at 1 year in the pre-CsA era, to 34% in the CsA era, to 9% in the tacrolimus era (P=0.008). The rate of technical failures (i.e., the risk of surgical complications) decreased from 30% in the pre-CsA era, to 14% in the CsA era, to 0% in the tacrolimus era (P=0.001). Patient survival rates at 1 year have ranged between 88% and 95% in the three eras (P=NS). Matching for at least one antigen on each HLA locus and avoiding HLA-B mismatches significantly decreased the incidence of rejection. The incidence of native kidney failure due to drug-induced toxicity decreased significantly over time, in part because only recipients with pretransplant creatinine clearance > or = 80 ml/min received transplants.

CONCLUSIONS

Solitary pancreas transplantation has become a viable alternative for nonuremic patients with labile IDDM. The risks of surgical complications and drug-induced nephrotoxicity have significantly decreased over time. Using tacrolimus as the mainstay immunosuppressant, patient and graft survival rates now no longer trail those of simultaneous pancreas-kidney transplantation.