• Donor;
  • import;
  • pancreas;
  • survival;
  • transplant


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References

The shortage of deceased donor organs for solid organ transplantation continues to be an ongoing dilemma. One approach to increase the number of pancreas transplants is to share organs between procurement regions. To assess for the effects of organ importation, we reviewed the outcomes of 1014 patients undergoing deceased donor pancreas transplant at a single center. We performed univariate and multivariate analyses of the association of donor, recipient and surgical characteristics with patient outcomes. Organ importation had no effect on graft or recipient survival for recipients of solitary pancreas transplants. Similarly, there was no effect on technical failure rate, graft survival or long-term patient survival for simultaneous kidney–pancreas (SPK) recipients. In contrast, there was a significant and independent increased risk of death in the first year in SPK recipients of imported organs. SPK recipients had longer hospitalizations and increased hospital costs. This increased medical complexity may make these patients more susceptible to short-term complications resulting from the longer preservation times of import transplants. These findings support the continued use of organ sharing to reduce transplant wait times but highlight the importance of strategies to reduce organ preservation times.


cerebrovascular accident




donor specific antibody


end stage renal disease


mycophenolate mofetil


organ procurement organization


pancreas after kidney transplant


pancreas donor risk index


panel of reactive antibodies


pancreas transplant alone


simultaneous pancreas-kidney transplant


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References

Pancreas transplantation can increase longevity and quality of life (1–5). The number of deceased organ donors has plateaued whereas the number of patients waiting for a pancreas transplant has continued to rise (6,7). This discrepancy has resulted in an increasing gap between supply and demand, and in an increased transplant wait time (7). Usage of pancreata from consented organ donors is less than for other organs (6). The 72% nonrecovery rate of consented potential pancreas donors is at an all-time high (6). Strategies aimed at improving conversion rates of potential pancreas donors could increase the number of organs available for transplant.

One barrier to the increased usage of pancreata is the reliance on locally procured organs. Many centers limit importation of pancreata for transplant because of a perceived increase in adverse outcomes with these donors. This barrier is made more complex by the lack of uniform pancreas allocation guidelines. As we move toward a unified allocation scheme for pancreas and kidney allografts, it is essential to define any increased risk with organ importation. For this reason we sought to compare the outcomes of transplanting locally procured versus imported pancreata.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References

Study population

We performed a retrospective review of all pancreas transplants performed at the University of Minnesota from July 1998 to June 2008. Approval was obtained from the institutional review board. Donor and recipient characteristics were obtained from electronic and paper medical records. Local donors were from the LifeSource (MNOP, St. Paul, MN, USA) organ procurement organization (OPO) territory that includes Minnesota, North Dakota, South Dakota and three counties in western Wisconsin.


Pancreas transplants were performed through midline laparotomy either as a pancreas transplant alone (PTA), pancreas after kidney transplant (PAK) or simultaneous kidney–pancreas transplant (SPK). The pancreaticoduodenal graft was placed intraabdominally. Systemic venous outflow was performed to iliac vessels or vena cava. Exocrine drainage was performed via bladder or enteric anastomosis. Postperfusion organ quality was assessed by the surgeon using a grading scheme (A–E) based on organ perfusion and the presence of edema, hemorrhage or pancreatitis. Patients were anticoagulated postoperatively with heparin and aspirin.


Immunosuppression consisted of induction with five to seven doses of an anti-T-cell antibody (equine antithymocyte globulin [ATGAM, Pfizer, New York, NY, USA] or rabbit antithymocyte globulin [Thymoglobulin, Genzyme, Cambridge, MA, USA]) and pulse steroids, followed by maintenance therapy with tacrolimus and mycophenolate mofetil (MMF). Before 2001, patients were maintained on low dose prednisone. In 2001, we converted to a steroid avoidance protocol and patients were rapidly tapered off of steroids after induction. Between 2003 and 2005, 134 patients were enrolled in a calcineurin inhibitor and steroid avoidance trial including induction therapy with alemtuzumab (Campath, Genzyme) and MMF maintenance. Anti-IL-2R antibody was used in addition to other induction agents 1998–1999 and then in selected patients.

Statistical analysis

Herein, we assess the association between local and imported pancreas transplants and five outcomes: technical failure, 1-year death-censored (DC) pancreas graft loss, 1-year mortality and long-term patient and graft survival. Time-to-event analysis was used to evaluate the relationship between organ offers and outcomes, adjusting for potential confounders that were present. Hospital and transplant outcomes including length of stay, hospital cost (year adjusted to 2008 dollars), pancreatic rejection, technical failure, graft failure and death were recorded. Donor and recipient characteristics, procurement details, surgical technique and organ quality were analyzed and correlated with recipient outcomes using univariate and multivariate analyses.

Donor risk factors examined included the following: donor age, body mass index (BMI), elevated recent/terminal creatinine (Cr), preservation time, mechanism of death, location of procurement and history of pancreatitis, drug or alcohol abuse (as defined by the reporting OPO). Recipient factors were age, gender, pancreas retransplant and pretransplant vascular disease including myocardial infarction, coronary artery bypass grafting, percutaneous coronary intervention, stroke, transient ischemic attack, claudication, amputation or peripheral bypass grafting/intervention. Surgical factors were type of transplant, bladder/enteric drainage and reperfusion grade. Pancreas donor risk index (pDRI) was calculated as described (8). Immunologic risk factors included donor/recipient HLA matching and panel of reactive antibody (PRA) values (peak class I, class II and cPRA). Listing parameters were interval from listing to transplant and position on match run list at organ offer. Immunosuppression induction and maintenance agents were included.

Statistical analysis was performed using version 9.2 of SAS (SAS Institute Inc., Cary, NC, USA). Univariate statistics were summarized as a count and percentage or a mean and standard error of the mean (SEM). For bivariate associations between risk factors and outcomes, categorical variables were tested using the chi-square statistic. The t-test was applied to continuous variables. Graft and patient survival was assessed by the Kaplan–Meier life table method using log-rank and Wilcoxon test statistics (9,10). Confounding risk factors statistically related to the outcomes and organ offers were incorporated into the multivariate analysis. Factors associated with the outcomes were selected for inclusion in separate Cox proportional hazards regression models for each outcome in each of the three transplant types. The proportionality assumption was tested by comparing log–log curves (11). The independent association of organ offers and outcomes were adjusted for confounding factors.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References

During the 10-year period from July 1998 to June 2008, we performed 1079 pancreas transplants. One thousand fourteen transplants were included in this analysis. The remaining 19 living-donor pancreas transplants, 37 simultaneous living donor kidney and deceased donor pancreas and 9 transplants with insufficient donor or recipient records were excluded. Of the analyzed transplants, 264 were SPK transplants, 298 were PTA and 452 were PAK. Seven hundred five transplants (70%) were imports and 309 (30%) were from local donors. Import donors were used relatively more frequently for PTA (77% import) and PAK (77% import) transplants whereas SPK (47% import) transplants were more likely to use local donors. Organs were imported from across the United States, including several donors from Alaska and Hawaii (Figure 1).


Figure 1. Geographic distribution of origin for imported pancreas organ donors. The donor hospital location is depicted by regional density clustering ( Larger circle size represents more donors and the number of donors from that region is indicated. Breakdown of category of transplant is noted by the colorized legend.

Download figure to PowerPoint

Donor and recipient characteristics were comparable between import and local organs (Table 1). Local and import donors were of similar age, BMI, terminal Cr (1.03 ± 0.04 vs. 1.15 ± 0.04, p = 0.090), and gender (58% vs. 61% male, p = 0.431). The number of donors with age >50, BMI ≥ 30 or terminal Cr ≥ 2.5 were not different. Import donors were more likely to be of non-Caucasian ethnicity (8% vs. 4%, p = 0.010). Mechanism of donor death differed between UNOS regions (p = 0.016) with a higher percentage of deaths from gunshot wounds in imports (14.9% vs. 6.6%, p = 0.040) and a reciprocal increase in stroke and other types of head trauma for local donors. There was no statistical difference between the groups in the frequency of donor drug or alcohol abuse.

Table 1.  Characteristics of 1014 local and import pancreas offers
 Organ offerp-Value
Local (n = 309)Import (n = 705)
Transplant type
Preservation time, mean ± SEM15.6 ± 0.3 18.8 ± 0.2 <0.001
Preservation time ≥ 20 hours6119.9%26738.4%<0.001
Donor characteristics
 Age, mean ± SEM31.0 ± 0.8 29.4 ± 0.5 0.079
 Aged 50 years or older299.4%436.1%0.060
 BMI, mean ± SEM24.4 ± 0.3 24.9 ± 0.2 0.097
 BMI ≥ 30 KG/m26019.5%15522.3%0.310
 Recent creatinine ≥2.5 mg/dL62.2%203.1%0.470
 Black race61.9%426.0%0.005
 Asian race10.3%91.3%0.157
 Cause of death - CVA10734.6%22331.8%0.370
 Donation after cardiac death10.3%223.1%0.006
Surgery characteristics
 Bladder drainage18459.5%49069.5%0.002
 Reperfusion grade (C, D or E)3814.0%7512.8%0.620
Recipient characteristics
 Female recipient15349.5%35450.2%0.830
 Age recipient, mean ± SEM43.9 ± 0.5 43.1 ± 0.3 0.166
 Pretransplant vascular disease14145.6%29141.3%0.190

Recipient age, gender, percentage retransplants and percent with preexisting vascular diseases were not different between import and local transplants (Table 1). Exocrine drainage was more often by bladder drainage in imported organ transplants than in local organs (70% vs. 60%, p = 0.002) which corresponds with the increased frequency of imported organs for use in PTA and PAK transplants and our standard practice of bladder drainage for solitary pancreas transplants. Bladder drainage was performed more frequently in solitary pancreas transplants to use urinary amylase as an alternative to Cr as a secondary indicator of pancreas rejection. Organ quality as assessed by intraoperative evaluation of reperfusion grade was not different.

Cold preservation time was on average 3 h longer for imported organs (18.8 h vs. 15.6 h, Figure 2) and roughly correlated with distance away from the transplant center (Figure 3). There was a trend toward decreased graft survival beyond 20 h of cold ischemia but this comparison failed to reach statistical significance (p = 0.110, data not shown).


Figure 2. Comparison of preservation times for import and local organs for each transplant category. Error bar represents mean ± SEM.

Download figure to PowerPoint


Figure 3. Preservation time as a function of distance away from transplant center. Cold ischemia time for each import donor is plotted as function of linear distance between donor hospital and transplant center. Linear correlation (R2 value of 0.033) is indicated.

Download figure to PowerPoint

Subgroup analysis for each of the three pancreas transplant types showed that donor and recipient characteristics used in the calculation of the pDRI for local and import donors were very similar (Table 2). The only consistent difference was a longer cold ischemia time (CIT).

Table 2.  Comparison of donor characteristics according to pancreas transplant type
Recipient characteristics
 Age (yr)42.3 ± 1.241.2 ± 0.70.42843.2 ± 0.843.8 ± 0.40.46245.2 ± 0.744.3 ± 0.80.402
 Pre-existing vascular disease29.4%26.1%0.58052.0%48.9%0.58048.9%48.0%0.880
 Female recipient64.7%62.2%0.70048.0%45.1%0.60043.2%42.4%0.900
Donor characteristics:
 Age (yr)26.8 ± 1.528.1 ± 0.80.43629.8 ± 1.229.0 ± 0.70.59533.9 ± 1.233.1 ± 1.20.624
 Black race2.9%4.8%0.5103.9%6.6%0.3100.0%6.4%0.002
 Asian race0.0%1.7%0.2730.0%0.9%0.3470.7%1.6%0.500
 BMI23.9 ± 0.524.3 ± 0.30.43724.1 ± 0.425.0 ± 0.30.07524.8 ± 0.425.6 ± 0.50.218
 Height (cm)169.0 ± 2.1170.1 ± 1.30.689171.9 ± 1.0169.5 ± 1.20.273172.2 ± 1.0165.3 ± 3.00.020
 Cause of death: CVA27.9%30.0%0.74034.3%32.2%0.68038.1%34.1%0.500
 Pancreas preservation time (hr)13.9 ± 0.518.7 ± 0.3<0.001 15.6 ± 0.518.3 ± 0.2<0.001 16.5 ± 0.420.2 ± 0.3<0.001 
 Donation after cardiac death1.5%4.8%0.2200.0%1.7%0.1820.0%4.0%0.017
 Serum creatinine > 2.5 mg/dL3.8%2.9%0.7293.2%4.1%0.7090.8%0.8%0.991
Donor risk index1.247 ± 0.0571.363 ± 0.0370.1211.192 ± 0.0411.271 ± 0.0230.0991.546 ± 0.0601.655 ± 0.0680.231

Our surgical team procured the organs in 52% of local donor cases. Other surgical teams procured 98.5% of imported organs. Patient (p = 0.910) and DC graft survival (p = 0.754) were the same whether the organs were recovered by our surgical team or that from another center.

In the initial analysis, recipient outcomes were examined (Table 3, Figure 4). There were no statistically significant differences in patient survival, DC graft survival, overall graft survival or technical failure rate between local organs and imports (all p > 0.1).

Table 3.  Patient survival, graft survival and technical failure rates in first year after transplant
  1. Percent occurrence.

  2. 1Signifies a statistical difference in the rates (Wilcoxon/log-rank tests; p-values 0.003/0.003), all other p > 0.1.

Technical failure (90 day) 8.811.3 4.9 9.712.217.6 9.111.6
Death-censored pancreas graft survival (1 yr)74.674.782.079.982.882.980.778.6
Overall pancreas graft survival (1 yr)73.571.780.477.476.369.677.074.2
Patient survival (1 yr)

Figure 4. Patient and death -censored graft survival for local and import pancreas transplants. The survival of (A–D) patients and (E–F) pancreas grafts for each category of transplant ([A, E] overall pancreas; [B, F] PTA; [C, G] PAK and [D, H] SPK) was analyzed by Kaplan–Meier survival analysis. Statistical comparison values are shown (log-rank/Wilcoxon).

Download figure to PowerPoint

Subgroup analysis of SPK, PAK and PTA transplants demonstrated an increased 1-year patient mortality for SPK recipients receiving import organs (Tables 3 and 4). One-year patient survival in the local group was 95% compared with 83% for imports (p = 0.003). This increased first-year mortality was not seen for imported organs in PAK or PTA recipients. The overall 89% first-year patient survival for all SPK recipients was less than for PTA (96%, p < 0.001) or PAK (98%, p < 0.001) recipients. Relative risk of first-year SPK patient death was 3.7-fold greater for import organs (Table 4). The patient survival difference was only seen in SPK recipients and only during the first year after transplant. SPK recipients of import pancreata alive at 1 year had continued survival that was indistinguishable from local donor transplants (p = 0.466). Long-term patient survival was equivalent between import and local organs for all groups (Figure 4).

Table 4.  Relative risk of technical failure, death-censored pancreas graft failure, pancreas graft failure and patient death at 1 year for import versus local pancreas offers
Relative risk95% CIp-ValueRelative risk95% CIp-ValueRelative risk95% CIp-Value
Technical failure (90 day)1.290.529, 3.1220.5802.020.790, 5.1660.1421.500.794, 2.8160.213
Death-censored pancreas graft failure1.010.587, 1.7370.9731.140.677, 1.9130.6261.020.546, 1.8990.953
Pancreas graft failure1.090.645, 1.8340.7521.180.725, 1.9350.4991.360.850, 2.1610.201
Patient death2.440.309, 19.2660.3970.820.165, 4.0420.8033.701.476, 9.2540.005

Pancreas importation had no statistically significant impact on graft survival for any of the transplant categories (Figure 4; Tables 3 and 4). There appeared to be an increase in technical failures with imported organs but this difference failed to achieve statistical significance (all p > 0.1).

For the multivariate analysis, the following outcomes were examined: patient mortality at 1-year, DC pancreas graft failure at 1-year and technical failure. Factors associated with these outcomes were included in separate models for each of the three transplant types. To simplify the model, donor related factors were compiled in the pDRI (Table 2). Inclusion of each of the parameters separately in the Cox model resulted in nearly identical results (not shown). The Cox model multivariate analysis demonstrated that outcomes with import pancreata were the same as for local donors for all outcomes except for first-year mortality in SPK recipients (Tables 5–7). Import organs for any of the transplant types were not associated with increased risk of technical failure (Table 5), pancreas graft loss (Table 6) or all-cause pancreas graft loss (not shown).

Table 5.  Cox proportional hazards regression model for the risk of technical failure after pancreas transplant
Relative risk95% CIp-ValueRelative risk95% CIp-ValueRelative risk95% CIp-Value
Source of donor
 Local1.00  1.00  1.00  
 Import1.180.470, 2.9590.7252.500.748, 8.3340.1370.840.400, 1.7660.647
Pancreas donor risk index0.990.476, 2.0750.9871.970.859, 4.4970.1102.121.352, 3.3220.001
 Reperfusion grade (C, D or E)5.672.439, 13.1800.0003.251.398, 7.5520.0063.261.490, 7.1370.003
 Bladder drainage0.390.156, 0.9760.0440.680.299, 1.5660.3690.460.192, 1.0930.079
Recipient characteristics
 Female recipient0.470.213, 1.0510.0660.490.210, 1.1280.0931.750.815, 3.7390.152
 Age recipient1.010.970, 1.0530.6140.990.939, 1.0370.5990.970.931, 1.0160.213
 Retransplant0.980.036, 2.7190.9742.000.930, 4.3140.0761.390.458, 4.2230.560
 Pretransplant vascular disease0.280.081, 0.9620.0430.970.439, 2.1380.9381.430.670, 3.0510.355
Table 6.  Cox proprotional hazards regression model for the death-censored pancreas graft failure after pancreas transplant
Relative risk95% CIp-ValueRelative risk95% CIp-ValueRelative risk95% CIp-Value
Source of donor
 Local1.00  1.00  1.00  
 Import0.980.557, 1.7240.9431.370.726, 2.5850.3310.580.272, 1.2350.158
Pancreas donor risk index1.280.807, 2.0400.2932.121.258, 3.5580.0052.151.340, 3.4400.001
 Reperfusion grade (C, D or E)2.291.177, 4.4540.0152.541.394, 4.6130.0022.731.231, 6.0340.013
 Bladder drainage0.600.334, 1.0930.0960.870.478, 1.5810.6460.480.203, 1.1100.085
Recipient characteristics
 Female recipient1.110.661, 1.8710.6880.870.515, 1.4610.5941.470.692, 3.1090.318
 Age recipient0.970.946, 0.9970.0270.980.945, 1.0120.1990.980.934, 1.0180.251
 Retransplant1.881.093, 3.2430.0232.171.305, 3.5890.0032.470.932, 6.5380.069
 Pretransplant vascular disease0.690.358, 1.3360.2721.370.807, 2.3200.2451.250.593, 2.6210.561
Table 7.  Cox proprotional hazards regression model for first-year patient mortality after SPK pancreas transplant
Relative risk95% CIp-Value
Source of donor
 Import4.921.717, 14.0880.003
Pancreas donor risk index1.590.967, 2.5980.067
 Reperfusion grade (C, D or E)0.980.360, 2.6590.965
 Bladder drainage1.380.480, 3.9620.551
Recipient characteristics
 Female recipient0.140.036, 0.5220.004
 Age recipient1.010.958, 1.0700.667
 Pretransplant vascular disease1.380.513, 3.7070.524
 Peak PRA ≥ 80%4.541.301, 15.8770.018

In SPK transplants, import recipients were more than four times more likely to die in the year after pancreas transplant than recipients receiving local organs (Table 7). The source of donor was unrelated to mortality for PTAs and PAKs and there were too few deaths in these transplants for the Cox model to converge.

To investigate the basis for the increased difference in first-year patient survival for import SPK recipients, we examined this group in more detail. We sought to determine if there were differences between import and local pancreas recipients or if there were differences between SPK recipients and solitary organ transplants that might make SPK imports more susceptible to adverse health events leading to early death.

First, we examined the cause of death in these patients (Figure 5). Infections were the most common cause of first-year death in SPK recipients. These included two bacterial infections (peripancreatic infections progressing to sepsis and multiorgan failure), three fungal infections (Aspergillus, Mucormycosis, disseminated Candida infection) and one viral infection (West Nile virus). Several deaths were technical in origin. One developed cardiac arrest after a postoperative bleed. Another had a respiratory arrest after a significant bleed from a rupture of a subcapsular hematoma after a kidney biopsy. A third died of multiorgan failure after a technically challenging transplant resulted in an ischemic limb. Other causes of death included two cases of hepatic failure, one suicide and one refusal of medical care. Susceptibility to individual causes of death in recipients of import organs failed to achieve statistical significance, likely due to low overall patient numbers in individual categories.


Figure 5. Causes of early death in transplant recipients. Numbers of deaths in the first year for PTA, PAK and SPK primary pancreas transplant recipients due to each category of medical condition are reported.

Download figure to PowerPoint

Next, we examined SPK transplants in more detail to determine if there were other confounding factors that might explain the increased mortality (Table 8). We looked at the measures of immunologic risk including number of HLA mismatches, percent with zero-mismatch transplants, peak class I, class II, cPRA and overall PRA values, number with PRA of 0% and number with PRA ≥80%. Each of these factors was added sequentially into the multivariate model. Only zero-mismatch (p = 0.042) and PRA ≥ 80% (p = 0.018) achieved statistical significance and only the latter improved the overall Cox model. None of these factors removed organ importation from significance. Donor specific antibody (DSA) testing only began after 2006 when single antigen bead assays were available. We did not transplant patients with known DSA.

Table 8.  Comparison of listing, immune risk, immunosuppression and hospitalization parameters
  1. 1Maximum PRA calculated by any assay.

  2. 2Cost adjusted to 2008 dollars.

Transplant listing
 Interval, listing to transplant (days)287.4 ± 116.1300.6 ± 71.30.934286.9 ± 129.1326.3 ± 50.50.730833.0 ± 116.3397.8 ± 84.40.011
Sensitization/immune matching
 Number HLA-A, -B, -DR mismatches2.75 ± 0.1482.74 ± 0.0790.9693.52 ± 0.1233.23 ± 0.0690.0453.60 ± 0.1332.68 ± 0.183<0.001 
 0 mismatch4.5%3.6%0.7380.0%1.2%0.2755.0%28.2%<0.001 
 PRA = 0%22.2%58.1%0.05866.7%54.8%0.28062.1%68.8%0.520
 Peak PRA ≥ 80%123.9%10.0%0.00312.0%6.0%0.04419.6%13.0%0.150
 Peak PRA class I (%)16.8 ± 4.07.5 ± 1.40.00611.0 ± 2.74.8 ± 0.90.0048.6 ± 2.48.3 ± 2.60.936
 Peak PRA class II (%)16.6 ± 4.36.9 ± 1.50.0088.6 ± 2.53.8 ± 0.90.0288.0 ± 2.27.2 ± 2.50.810
 Peak overall PRA (%) (1)31.3 ± 5.017.3 ± 2.10.00323.4 ± 3.513.1 ± 1.40.00224.9 ± 3.215.8 ± 2.90.040
 IL-2R Antibody50%59%0.19061%63%0.69057%62%0.380
 Discharge on steroids47%54%0.32062%67%0.42066%54%0.060
Hospitalizations and costs
 Length of stay, transplant admission (days)11.3 ± 0.911.7 ± 0.50.71110.1 ± 0.611.2 ± 0.60.31618.9 ± 1.619.9 ± 1.80.678
 Length of stay, year one (days)35.3 ± 3.430.7 ± 1.50.17429.7 ± 2.030.7 ± 1.30.70042.3 ± 2.941.4 ± 3.20.830
 Hospital costs, transplant admission, $2$118716 ± 4006$117567 ± 20610.793$108492 ± 2630$113218 ± 24460.323$208007 ± 11294$206137 ± 88310.895
 Hospital costs, year 1, $2$163551 ± 10042$147166 ± 40720.078$136897 ± 4768$149380 ± 43130.140$250679 ± 14132$242482 ± 113680.648

We evaluated induction and maintenance immunosuppression for significance in the Cox model. None of these regimens, including use of alemtuzumab (p = 0.537), maintenance steroids (p = 0.635), anti-IL2R (p = 0.895) or individual maintenance agents were significant in the model. Pancreas rejection in the first year did not predict mortality (p = 0.493). We looked for the presence of confounding era effects. The number of import organs used throughout the 10-year period of the study ranged from 59% to 81%, but individual cohort years were not associated with increased mortality. Similarly, eras defined by quartiles of sequential transplants or by immunosuppressive regimen were not significant (all p > 0.2).

Because we could not identify any confounding factors that explained the apparent increase in mortality in import SPK recipients, we looked for differences between SPK transplants and solitary organ transplants that might indicate an increase in medical complexity or susceptibility to complications after transplant (Table 9). The waiting time for SPK recipients was more than twice that of PTA or PAK recipients. With the local organ allocation policies during this study, the waiting time for locally procured SPK transplants was even greater (833 ± 116 days), prompting an increased reliance on imported organs and a usage of organs of greater overall risk. This is supported by an increased mean pDRI for SPK transplants (pDRI = 1.60) compared with PTA (1.34) or PAK (1.25) organs. SPK transplants had the greatest preservation time of the three groups and import SPK transplants had the highest of any category (Figure 2).

Table 9.  Transplant waiting times, length of stay and hospital costs
Mean ± SEMMean ± SEMp-ValueMean ± SEMp-Value
Transplant waiting time (days)661.6 ± 85.6298.1 ± 61.20.018315.6 ± 50.10.009
Length of stay transplant admit (days)19.4 ± 1.211.6 ± 0.4<0.00111.0 ± 0.5<0.001
Length of stay year 1 (days)41.9 ± 2.231.7 ± 1.4<0.00130.4 ± 1.1<0.001
Donor risk index1.60 ± 0.051.34 ± 0.03<0.0011.25 ± 0.02<0.001
Hospital costs transplant admit (2008$)$206983 ± 7017$117831 ± 1831<0.001$112163 ± 1990<0.001
Hospital costs year 1 (2008$)$246192 ± 8905$150923 ± 3906<0.001$146594 ± 3524<0.001

To assess for greater postoperative morbidity and complexity in SPK recipients, we compared hospital length of stay and hospital costs during the transplant admission and for the first year after transplant. These function as surrogate indicators for in-hospital medical complexity. Length of stay was nearly twice as long for SPK transplant admissions and ∼40% greater overall during the first year. Hospital costs were nearly double for SPK transplants. The waiting time for SPK transplants is longer than for other transplants, which prompted a reliance on imported organs with longer preservation time and greater pDRI values. Although these factors did not shorten graft survival, they did portend greater in-hospital medical complexity that likely resulted in the increased first-year mortality.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References

Here, we report on a series of 1014 deceased donor pancreas transplants performed at a single institution over the course of 1 decade. We have found no statistical difference in graft survival or technical failure rate for SPK, PTA or PAK pancreas transplants between local or imported pancreas donors. Overall patient survival was equivalent for pancreas recipients in general but there was an increased 1-year patient mortality for SPK recipients receiving import organs. The mortality difference was only present in the first year after transplant. Long-term patient survival for SPK recipients was equivalent for imports and local organs. PTA and PAK recipients had equivalent short- and long-term patient survival with imported organs.

The basis for the increased first-year mortality in import SPK recipients is not entirely clear. The increase was verified in a multivariate Cox model analysis. Donor, recipient and operative characteristics were very similar between import and local pancreata. The only clear differences were an increased preservation time and a corresponding higher pDRI. We performed a thorough search for confounding characteristics that might explain these unexpected findings. We found that recipient male gender and high PRA were risk factors for SPK recipient death but these parameters did not drop importation from significance. We looked extensively for other factors including immunosuppressive regimen, donor and recipient demographics, measures of immunologic risk, era effects and pancreas rejection episodes. None of these factors eliminated the impact of organ importation.

Due to past differences in organ allocation policies between SPK and solitary pancreata, our donor selection was less stringent for SPK transplants than solitary pancreas transplants. During the period of this study, combined kidney–pancreas placement in our OPO followed priority established on the kidney waiting list rather than prioritizing SPK transplants. This policy resulted in both increased wait times for SPK patients compared to solitary pancreas recipients and in a practice of accepting organs of potentially lesser quality due to the increased risk of death on the waitlist for SPK patients. Waiting times for SPK organs were twice that of solitary pancreata. This trend was also true for SPK imports where exporting OPOs favored retaining kidneys within their region and sharing pancreata as solitary organs only. Given the relatively poor performance of diabetic patients on dialysis, our practice was to be more inclusive in our acceptance of available organs for SPK transplants. Solitary organ recipients had the dual benefits of both improved overall health and reduced wait times. These patients were better able to wait for more ideal donors. This is evidenced by an increased donor age for SPK recipients compared to PTA or PAK recipients (p < 0.001) and an increase in the pDRI for imported organs.

There are several potential reasons that the apparent increase in first-year mortality was seen only in SPK transplants. SPK patients in general have greater medical comorbidity at the time of transplant. SPK and PAK are older and have a greater amount of pretransplant vascular disease than PTA recipients. PAK patients have already undergone a successful kidney transplant and thus have a selection bias for patients tolerating the initial operation and the benefit of a functioning kidney at the time of transplant. This medical complexity is supported by hospital lengths of stay for SPK recipients that are twice that of solitary pancreas transplants and by markedly greater hospital costs. The combination of more pretransplant comorbidity and somewhat lower donor quality likely predisposes SPK recipients to difficulties in the first year. These factors may increase the health impact of postoperative infection, cardiovascular insult or reperfusion pancreatitis. These complications may be more common or more serious in import organ recipients where the preservation times are longer.

The reduced patient survival in SPK transplants may be acceptable given the overall poor survival of diabetics on dialysis. One-year survival of incident end-stage renal disease (ESRD) patients due to diabetes mellitus was only 77–81% over the years of this study (12). Even with the apparent increased first-year mortality of import SPK recipients, the survival with transplant still exceeds that of patients who continue on dialysis. The benefit of SPK transplant is markedly evident beyond 1 year after transplant where patient survival stabilizes considerably. Five-year patient survival of SPK recipients in our patient cohort is >75% whereas 5-year survival of incident hemodialysis patients with diabetes is <31% (12).

Our results parallel and confirm the findings from smaller case series from other centers. One group reported 247 pancreas transplants of which 26% were imported (13). They found no difference in thrombosis rate, graft survival or patient survival with import pancreas transplants, nor did they find a difference in graft survival if the organs were procured by their surgeons. We similarly found no difference in outcomes if our surgical team or a surgical team from another center procured the organ. Of note, this group transported import pancreata by charter flight, which significantly reduces cold preservation time. The CIT for local donors was 8.2 h and importation only increased this time to 9.5 h. Such practice has been suggested to effectively convert import organs into local donor equivalents (14). We found that at 20 h of CIT, there was a trend toward decreased graft survival but this did not reach statistical significance. Other groups have also reported little impact of moderate prolongation of CIT on graft or patient survival (15,16). The use of fixed wing charter flights to transplant imported pancreata significantly reduces cold ischemia time but increases overall cost (17). We do not routinely charter flights for import pancreata. This approach results in an average cold ischemia time at our center that is significantly longer than that reported by other groups.

In comparing the outcomes reported here to those of other centers, several interesting findings are revealed. Unadjusted 1-year patient and graft survival at high volume centers is 97.3% and 74.5% for PTA recipients and 98.7% and 83.5% for PAK recipients, respectively (18). These values are nearly identical to survival rates in this report. Our overall SPK patient survival is 89%, which is slightly lower than reported by the Organ Procurement and Transplantation Network (OPTN) report or by other large series or high volume centers (13,19–21). Examination of the OPTN report confirms worse outcomes with longer preservation time, older recipients, retransplantation and older donors, all of which are highly represented in our transplant population. Approximately 80% of the pancreas grafts that we import have been turned down by other centers. We are inclusive in selection of both recipients and donors with advanced age and comorbidity. When comparing outcomes for our patients as stratified by pDRI, our results coincide closely with the survival curves reported in the description of this parameter (8).

This study confirms and updates our earlier report of pancreas transplants performed during previous eras at our institution wherein we similarly found little, if any, difference between locally procured or imported organs (22). This study overlaps the last era of the original report and studies the affect of organ importation in recent practice. Our results support the continued use of imported pancreata for transplant and demonstrate the importance of improved policies for organ allocation that enable increased usage of potential organs.

Policy changes that favor sharing of combined pancreas–kidney allografts will facilitate the improved accessibility of import organs. This study documents equivalent long-term patient and graft survival between imported organs and those procured locally, and supports the continued use of import organs for pancreas transplantation. The increased mortality observed in this study for import pancreas transplants was only in the first year and only for SPK transplants. This increased mortality is to a large part related to longer preservation times for import organs and is likely driven by local organ allocation and acceptance practices. This mortality may be acceptable given the poor survival of diabetics on dialysis. Strategies to reduce organ transport time including chartering of fixed-wing aircraft should be considered to minimize preservation time and maximize organ usage by sharing.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosure
  8. References
  • 1
    Becker BN, Brazy PC, Becker YT, et al. Simultaneous pancreas-kidney transplantation reduces excess mortality in type 1 diabetic patients with end-stage renal disease. Kidney Int 2000; 57: 21292135.
  • 2
    Gross CR, Limwattananon C, Matthees B, Zehrer JL, Savik K. Impact of transplantation on quality of life in patients with diabetes and renal dysfunction. Transplantation 2000; 70: 17361746.
  • 3
    Morath C, Zeier M, Döhler B, Schmidt J, Nawroth PP, Opelz G. Metabolic control improves long-term renal allograft and patient survival in type 1 diabetes. J Am Soc Nephrol 2008; 19: 15571563.
  • 4
    Nathan DM, Fogel H, Norman D, et al. Long-term metabolic and quality of life results with pancreatic/renal transplantation in insulin-dependent diabetes mellitus. Transplantation 1991; 52: 8591.
  • 5
    White SA, Shaw JA, Sutherland DER. Pancreas transplantation. Lancet 2009; 373: 18081817.
  • 6
    Tuttle-Newhall JE, Krishnan SM, Levy MF, McBride V, Orlowski JP, Sung RS. Organ donation and utilization in the United States: 1998–2007. Am J Transplant 2009; 9(Pt 2): 879893.
  • 7
    McCullough KP, Keith DS, Meyer KH, Stock PG, Brayman KL, Leichtman AB. Kidney and pancreas transplantation in the United States, 1998–2007: access for patients with diabetes and end-stage renal disease. Am J Transplant 2009; 9(Pt 2): 894906.
  • 8
    Axelrod DA, Sung RS, Meyer KH, Wolfe RA, Kaufman DB. Systematic evaluation of pancreas allograft quality, outcomes and geographic variation in utilization. Am J Transplant 2010; 10: 837845.
  • 9
    Kalbfleisch JD, Prentice RL. The Statistical Analysis of Failure Time Data. New York : Wiley; 1980.
  • 10
    Lee ET. Statistical methods for survival data analysis. Belmont , CA : Lifetime Learning Publications, 1980.
  • 11
    Allison PD, Institute SAS. Survival analysis using the SAS system: A practical guide. Cary , NC : SAS Institute; 1995.
  • 12
    USRDS. USRDS 2010 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases; 2010.
  • 13
    Fridell JA, Mangus RS, Hollinger EF, et al. No difference in transplant outcomes for local and import pancreas allografts. Transplantation 2009; 88: 723728.
  • 14
    Stratta RJ, Farney AC, Rogers J. Import pancreas allografts: Good from far or far from good? Transplantation 2009; 88: 622623.
  • 15
    Stratta RJ. Donor age, organ import, and cold ischemia: Effect on early outcomes after simultaneous kidney-pancreas transplantation. Transplant Proc 1997; 29: 32913292.
  • 16
    Sousa MG, Linhares MM, Gonzalez AM, et al. Multivariate analysis of risk factors for early loss of pancreas grafts among simultaneous pancreas-kidney transplants. Transplant Proc 2010; 42: 547551.
  • 17
    Orlowski JP, Jaynes CL, Spees EK. Practical reduction of transplantation costs. Use of commercial transportation instead of charter aircraft for sharing pancreas grafts. Arch Surg 1993; 128: 11111114.
  • 18
    SRTR. 2009 Annual Report of the U.S. Organ Procurement and Transplantation Network and the Scientific Registry of Transplant Recipients: Transplant Data 1999–2008. Rockville, MD; 2009.
  • 19
    Sollinger HW, Odorico JS, Becker YT, D’Alessandro AM, Pirsch JD. One thousand simultaneous pancreas-kidney transplants at a single center with 22-year follow-up. Ann Surg 2009; 250: 618630.
  • 20
    Kaufman DB, Leventhal JR, Gallon LG, Parker MA. Alemtuzumab induction and prednisone-free maintenance immunotherapy in simultaneous pancreas-kidney transplantation comparison with rabbit antithymocyte globulin induction—Long-term results. Am J Transplant 2006; 6: 331339.
  • 21
    Rajab A, Pelletier RP, Ferguson RM, Elkhammas EA, Bumgardner GL, Henry ML. Steroid-free maintenance immunosuppression with rapamune and low-dose neoral in pancreas transplant recipients. Transplantation 2007; 84: 11311137.
  • 22
    Sutherland DE, Gruessner RW, Dunn DL, et al. Lessons learned from more than 1000 pancreas transplants at a single institution. Ann Surg 2001; 233: 463501.