- Email: [email protected]
Prolonged preservation increases surgical complications after pancreas transplants
Prolonged preservation increases surgical complications after pancreas transplants Abhinav Humar, MD, Raja Kandaswamy, MD, Mary Beth Drangstveit, RN, Elizabeth Parr, RN, Angelika G. Gruessner, PhD, and David E.R. Sutherland, MD, Minneapolis, Minn
Background. Prolonged preservation of the donor organ may result in delayed graft function or nonfunction after most organ transplants. We studied whether or not prolonged preservation increases surgical complications after pancreas transplants. Methods. Between January 1, 1994, and September 30, 1998, a total of 294 recipients underwent cadaver pancreas transplants at our institution. Recipients were analyzed in 2 groups: those with pancreas preservation time ≤ 20 hours (n = 211) versus > 20 hours (n = 83). Results. Demographic data were similar between the 2 groups, except that mean donor age in the prolonged preservation group was significantly lower. Despite use of younger donors, prolonged preservation was associated with an increased incidence of surgical complications, most notably leaks, thrombosis, and wound infections. Grafts with prolonged preservation were more often noted by the transplant surgeon to be edematous after reperfusion, although the incidence of hyperamylasemia posttransplant did not differ between the 2 groups. Graft and patient survival rates also did not differ between the 2 groups. The incidence of early graft loss (< 3 months) was, however, higher in the prolonged preservation group (20.5% versus 9.0%, P = .04). Conclusions. Prolonged preservation of the donor organ increases the incidence of surgical complications after pancreas transplants. All attempts should be made to minimize preservation time, keeping it below 20 hours, if possible. (Surgery 2000;127:545-51) From the Department of Surgery, University of Minnesota, Minneapolis
THE INTRODUCTION OF EFFECTIVE preservation solutions, such as the University of Wisconsin solution, has allowed significant prolongation of cold ischemia for all solid organ grafts.1-3 Yet, eventually, organ damage ensues. Clinically, such damage may manifest itself with delayed graft function or nonfunction. The functional consequences of prolonged preservation have been well documented for kidney,4 liver,5 and pancreas6 transplants; but other consequences—such as surgical complications—have not been as fully described. Ischemic injury may lead to cellular swelling and graft edema,7 which may then lead to surgical complications with significant morbidity. The purpose of this study was to document the effects of prolonged preservation on the incidence of surgical complications after pancreas transplants. Attempts to shorten preservation time Accepted for publication November 25, 1999. Reprint requests: Abhi Humar, MD, Department of Surgery, University of Minnesota Medical School, Box 328, Mayo Building, 420 Delaware St S.E., Minneapolis, MN 55455. Copyright © 2000 by Mosby, Inc. 0039-6060/2000/$12.00 + 0 doi:10.1067/msy.2000.104742
11/56/104742
could be more aggressive if the impact on outcome were known to be significant. METHODS Patients. Between January 1, 1994, and September 30, 1998, a total of 294 recipients underwent cadaver pancreas transplants at the University of Minnesota. Recipients were analyzed in 2 groups: those with pancreas preservation time ≤ 20 hours (n = 211) versus > 20 hours (n = 83). Mean preservation time in the first group was 14.6 hours versus 22.8 hours in the latter group (P = .01). Transplant categories (simultaneous pancreas-kidney [SPK], pancreas after kidney [PAK], and pancreas transplant alone [PTA]) were proportionately similar between the 2 groups. Surgery. The donor and recipient operations, and selection criteria, have been detailed previously.8-10 Organs were flushed in situ with University of Wisconsin solution and preserved in cold University of Wisconsin solution on ice. For locally procured organs, the donor duodenum was flushed with an antibiotic solution before being stapled shut. In the recipient operation, the pancreas was anastomosed to the iliac vessels (on the right side if possible) after reconstruction of the SURGERY 545
546 Humar et al
donor superior mesenteric and splenic arteries with a donor iliac artery Y-graft. Exocrine secretions were managed with a duodenocystostomy or duodenoenterostomy. Intraoperative data on the appearance of the transplanted pancreas were noted by the attending surgeon and later entered into a database. Immediately after reperfusion, the consistency of the pancreas was recorded as being soft, firm, or hard. The presence or absence of edema in the pancreas, as assessed by the transplant surgeon, was also noted. Postoperative care. All recipients received induction therapy with an antilymphocyte preparation. Maintenance immunosuppression was initiated with tacrolimus, mycophenolate mofetil, and prednisone. During the first 6 months posttransplant, tacrolimus doses were adjusted to maintain serum levels of 10 to 15 ng/dL, then 8 to 10 ng/dL thereafter. Recipients received antiviral prophylaxis (acyclovir or ganciclovir) and antibacterial prophylaxis with a broad-spectrum antibiotic (imipenem, ampicillin/sulbactam, or ceftazidime). Fluconazole was also given during the first week posttransplant for antifungal prophylaxis. Most recipients also received standard prophylaxis against vascular thrombosis: low-dose heparin (300 to 500 units/hr) for the first 5 days posttransplant and acetylsalicylic acid (325 mg/d) for the first 3 months posttransplant. Recipients with a bleeding complication or any contraindication to anticoagulation did not receive low-dose heparin. Surgical complications. Surgical complications during the first 3 months posttransplant were included in this analysis.11,12 Complications included relaparotomy, vascular graft thrombosis, intraabdominal infections, bleeding, pancreatitis, and anastomotic or duodenal stump leaks. Relaparotomy was defined as any reoperative procedure involving the intraperitoneal or retroperitoneal space during the first 3 months posttransplant. Thrombosis was diagnosed by clinical symptoms, imaging study results, intraoperative findings, or a combination thereof. Thrombosis included venous and arterial thrombosis. Bleeding was defined as a significant complication if it necessitated a reoperation or a transfusion of more than 2 units of red blood cells. For all intra-abdominal infections included in this study, patients were symptomatic, with documented positive culture results. Cultures were obtained from aspirations guided by computed tomography or ultrasonography, or at the time of abdominal exploration. Anastomotic or duodenal stump leaks were diagnosed by clinical symptoms, imaging study results, laboratory findings, or a combination thereof.
Surgery May 2000
Postoperative pancreatitis was defined by a serum amylase level above 120 mg/dL for the first 5 postoperative days, accompanied by pain and tenderness over the graft. Rejection. Pancreas graft rejection after a bladder-drained transplant was suspected by a decrease in urinary amylase levels of 25% or more from baseline on 2 consecutive measurements and confirmed by biopsy in the majority of cases.13 Pancreas graft rejection after SPK transplants was frequently diagnosed by an increase in serum creatinine level; kidney biopsies were obtained whenever kidney graft rejection was clinically suspected. Treatment was generally with a course of antilymphocyte therapy. Data analysis. Categorical variables were analyzed by using the chi-square test, and when applicable, Fisher’s exact test. Continuous variables were analyzed parametrically with the Student t test. Pancreas graft and patient survival rates were calculated according to the Kaplan-Meier method. Pancreas graft loss was defined by the recipient’s return to exogenous insulin use after insulin independence. Calculation of patient survival included deaths occurring after pancreas and kidney graft loss. Survival rates were compared among groups with the generalized Wilcoxon’s test. A P value of ≤ .05 was considered significant. RESULTS Of 294 recipients, 211 received organs with a preservation time of ≤ 20 hours while 83 received organs with a preservation time of > 20 hours. Characteristics of the 2 groups are shown in Table I. The mean recipient age was the same in both groups, but the mean donor age was significantly lower in the prolonged preservation group (22.3 years versus 28.1 years, P = .04). The composition of the 2 groups by transplant category was similar, with SPK and PAK each accounting for roughly 40% of the transplants. Enteric drainage was performed in about 10% of recipients in both groups. Donor hemodynamic stability before organ procurement (as measured by the need for inotropes in the donor) was the same for both groups. Other characteristics of the donor, such as cause of death and renal function, were not significantly different between the 2 groups. Organs with lower preservation times were more likely to have been procured locally: 50.0% of organs with a preservation time of ≤ 20 hours were procured locally versus 35.4% of those with a preservation time of > 20 hours (P = .02). Early (< 3 months posttransplant) surgical complications for the 2 groups are shown in Table II.
Humar et al 547
Surgery Volume 127, Number 5 Table I. Characteristics of 2 groups Total number of recipients SPK/PAK/PTA Mean recipient age (y) Mean donor age (y) Mean preservation time (h) Enteric drainage: n (%) Donor inotrope use Cause of donor death Trauma CVA Donor creatinine (mg/dL) Local procurement
Preservation ≤ 20 h
Preservation > 20 h
P value
211 89/91/31 39.6 28.1 14.6 23 (10.9%) 56%
83 39/32/12 40.5 22.3 22.8 9 (10.8%) 58%
— NS NS .04 .01 NS NS
58% 25% 1.6 50.0%
64% 20% 1.5 35.4%
NS NS NS .02
SPK, Simultaneous pancreas-kidney; PAK, pancreas after kidney; PTA, pancreas transplant alone; CVA, cerebrovascular accident.
Table II. Surgical complications and outcomes Preservation ≤ 20 h Total number of recipients Complications Intra-abdominal infections Wound infections Bleeding Duodenal leaks Thrombosis Pancreatitis Relaparotomy rate Outcomes Acute rejection (3 m) Graft survival (3 m) Graft survival (3 y) Patient survival (3 y)
211
Preservation > 20 h 83
P value —
5.7% 10.4% 18.0% 5.2% 5.2% 11.4% 16.4%
8.4% 18.1% 18.0% 14.5% 10.8% 13.3% 19.3%
NS .07 NS .008 .08 NS .23
23.7% 91.0% 69.8% 92.1%
21.7% 79.5% 72.3% 93.5%
NS .04 NS NS
Prolonged preservation was associated with an increased likelihood for surgical complications, most notably leaks, thrombosis, and wound infections. The most dramatic difference in the 2 groups was in the incidence of duodenal leaks: 14.5% for preservation > 20 hours versus 5.2% for preservation ≤ 20 hours (P = .008). The leak rate increased proportionately with the preservation time (Fig 1). With a preservation time of < 15 hours, the leak rate was 1.4%. When the preservation time increased to 15 to 20 hours, the leak rate increased to 5.8% (P = .14); with 20 to 25 hours, 14.1% (P = .001); with 25 to 30 hours, 16.7% (P = .03). Though not as striking, a similar trend was seen when thrombosis was used as the dependent variable (Fig 2). The incidence of thrombosis was higher with preservation times of 20 to 25 hours and 25 to 30 hours, versus < 15 hours or 15 to 20 hours (P = .10). As a result of these complications, recipients in the prolonged preservation group tended to have a
higher relaparotomy rate during the first 3 months posttransplant (19.3% versus 16.4%, P = .23), but this trend was not statistically significant. The incidence of bleeding was the same in both groups. Grafts with prolonged preservation were more often noted to be edematous immediately after reperfusion, but this was a subjective evaluation by the transplant surgeon. The incidence of clinical pancreatitis, however, was not higher in the prolonged preservation group. Prolonged preservation did not have a significant impact on patient or graft (Fig 3) survival rates. At 3 years posttransplant, we noted no significant difference in either graft or patient survival rates between the 2 groups. However, the incidence of early graft loss (by 3 months posttransplant) was significantly higher with prolonged preservation. By 3 months posttransplant, 9.0% of grafts in the shorter preservation group had been lost versus 20.5% in the prolonged preservation group (P = .04). Most of
548 Humar et al
Surgery May 2000
Fig 1. Incidence of leaks versus preservation time. Total number of recipients in each group is shown. Asterisk, P < .05 versus incidence of leaks in < 15 hours preservation group.
Fig 2. Incidence of thrombosis versus preservation time. Total number of recipients in each group is shown.
these early graft losses represent failures secondary to technical complications, such as thrombosis or leaks, that lead to graft removal. DISCUSSION Pancreas transplants are still associated with the highest surgical complication rate of all routinely performed solid organ transplants.14 Several risk
factors have been identified that predict an increased likelihood for surgical complications, including older donor age, older recipient age, enteric (versus bladder) drainage, graft pancreatitis, and retransplants.11,12,14 Prolonged preservation and poor pancreas function have been associated,6,15 but the impact on surgical complications has not been fully described. This impact is impor-
Surgery Volume 127, Number 5
Humar et al 549
Fig 3. Pancreas graft survival rates.
tant to document: If it is significant, more aggressive efforts could be made to decrease preservation time. Previous reports have shown some influence of prolonged preservation on surgical complications such as thrombosis.16 In an earlier study from our center, logistic regression analysis revealed prolonged preservation to be a significant risk factor for death in PTA recipients.14 Our current study also suggests that prolonged preservation is a risk factor for surgical complications after pancreas transplants. Our 2 groups were demographically similar. Their composition by transplant category was nearly identical, with SPK and PAK each accounting for roughly 40% of the transplants in both groups. This fact is important to note, because the surgical complication rate is generally higher after a double-organ transplant than after single-organ transplants (PAK or PTA). The methods of exocrine drainage may also influence the complication rate. Enteric drainage and bladder drainage have slightly different complication rates. We have recently started to use enteric drainage more frequently, especially with SPK transplants where the serum creatinine level can be used to monitor for rejection. In this study, the percentage of recipients with enteric drainage was nearly identical for both groups. Donor factors may also influence surgical complications. Unstable donors may have organs with more ischemic damage. Using inotrope administration and mean serum creatinine in the donor as a crude approximation of donor stability, we were
not able to identify a significant difference in our 2 groups. Pancreases with prolonged preservation more commonly were imported from an outside region. This factor may have some impact: experience of the procuring team may have an impact on the surgical team posttransplant. Organs damaged at the time of procurement may lead to increased complications after the transplant. It is possible that the higher complication rate seen in the prolonged preservation group may be in part related to the higher percentage of imported organs in that group. However, as the number of centers performing pancreas transplants has dramatically increased, so has the number of surgeons capable of safely procuring a pancreas for transplant. Another difference between the 2 groups was that donors in the prolonged preservation group were significantly younger than in the other group. Therefore, a longer preservation time was accepted with younger donors. Earlier analyses from our center have consistently demonstrated that older donor age is a significant risk factor for surgical complications after pancreas transplants.11,12 But, even with younger donors, our current study shows that prolonged preservation was associated with an increased risk for surgical complications. Leaks, thrombosis, and wound infections were overall more common in the prolonged preservation group. These surgical complications, all associated with significant morbidity, accounted for the higher reoperation rate in the prolonged preservation group. For certain complications, such as leaks, there was very good correlation between
550 Humar et al
increasing preservation times and increasing incidence of the complication (Fig 1). This illustrates the importance of trying to minimize preservation times, rather than simply targeting for a maximum allowable time limit. The reasons for a higher surgical complication rate after prolonged preservation are unclear, but may in part be related to increased graft edema after prolonged preservation. Reperfusion after cold preservation results in some tissue injury, which manifests as graft edema and swelling.7,17 The significantly higher leak rate with prolonged preservation may also be attributable to a different sensitivity of the donor duodenum (versus the pancreatic parenchyma) to prolonged preservation. Data from the small bowel literature suggest that the bowel has a poor tolerance to prolonged preservation.18,19 Since the vast majority of pancreas transplants are performed with the donor duodenum intact, this may explain the higher incidence of leaks with increasing preservation times. We did not examine the effect of prolonged preservation on graft endocrine function, because our purpose was to examine its impact on surgical complications. In previous analyses from our center, preservation for more than 24 hours was possible, but did result in a trend toward slightly inferior endocrine function; a higher number of such recipients had abnormal oral glucose tolerance tests postoperatively.6,20-22 Other studies have also demonstrated decreased insulin secretion from organs preserved for more than 24 hours.15,23 We were not able to demonstrate a significant detrimental effect of prolonged preservation on long-term (ie, 3-year) graft survival. However, the 3month graft survival rate in the prolonged preservation group was significantly lower. These early graft losses reflect the higher surgical complication rate in this group and the consequence of adverse events such as thrombosis or leaks. In conclusion, preservation time has an impact on the development of surgical complications after pancreas transplants. Prolonged preservation, likely through the development of graft edema, leads to an increased incidence of complications, such as thrombosis and leaks. The incidence of surgical complications after pancreas transplants remains high, so all efforts must be made to minimize preservation time (preferably to < 20 hours). Special thanks to Mary Knatterud and Ernestine Draper for their invaluable help in the preparation of this manuscript.
Surgery May 2000 REFERENCES 1. Belzer FO, Southard JH. Principles of solid-organ preservation by cold storage. Transplantation 1988;45:673-6. 2. D’Alessandro AM, Stratta RJ, Sollinger HW, et al. Use of UW solution in pancreas transplantation. Diabetes 1989;38:7-9. 3. Abouna GM, Sutherland DER, Florack G, et al. Function of transplanted human pancreatic allografts after preservation in cold storage for 6 to 26 hours. Transplantation 1987;43:630-6. 4. Koning OH, Ploeg RJ, van Bockel JH, et al. Risk factors for delayed graft function in cadaveric kidney transplantation: a prospective study of renal function and graft survival after preservation with University of Wisconsin solution in multiorgan donors. Transplantation 1997,63:1620-8. 5. Porte RJ, Ploey RJ, Hansen B, et al. Long-term graft survival after liver transplantation in the UW era: late effects of cold ischemia and primary dysfunction. Transplant Int 1998,11:S164-7. 6. Morel P, Moudry-Munns K, Najarian JS, et al. Influence of preservation time on outcome and metabolic function of bladder-drained pancreas transplants. Transplantation 1990,49:294-303. 7. Schaapherder AF, de Roos A, Shaw PC, et al. The role of early baseline computed tomography in the interpretation of morphological changes after kidney-pancreas transplantation. Transplant Int 1993,6:270-6. 8. Gruessner RWG, Sutherland DER. Pancreas transplantation: Part I: The donor operation. Surg Rounds 1994;17:311-24. 9. Gruessner RWG, Sutherland DER. Pancreas transplantation: Part II: The recipient operation. Surg Rounds 1994; 17:383-91. 10. Gruessner RWG, Dunn DL, Gruessner AC, et al. Recipient risk factors have an impact on technical failure and patient and graft survival rates in bladder-drained pancreas transplants. Transplantation 1994;57:1598-1606. 11. Troppmann C, Gruessner AC, Sutherland DER, et al. Surgical complications requiring early relaparotomy after pancreas transplantation: a multivariate risk factor and economic impact analysis. Ann Surg 1998;227:255-68. 12. Troppmann C, Gruessner AC, Benedetti E, et al. Vascular graft thrombosis after pancreatic transplantation: univariate and multivariate operative and nonoperative risk factor analysis. J Am Coll Surg 1996;182:285-316. 13. Gruessner RWG, Sutherland DER. Clinical diagnosis in pancreatic allograft rejection. In: Solez K, Racusen LC, Billingham ME, editors. Solid organ transplant rejection: mechanism, pathology, and diagnosis. New York: Marcel Dekker, Inc; 1996. p. 455-89. 14. Gruessner RWG, Sutherland ER, Troppmann C, et al. The surgical risk of pancreas transplantation in the cyclosporine era: an overview. J Am Coll Surg 1997;185:128-43. 15. Lopez-Medrano RM, Ponce J, Varela G, et al. Extensive decrease in insulin secretion by the pancreas preserved for 24 hours in UW-1 solution. Transplant Proc 1991;23:1676-8. 16. Sutherland DE. Report from the International Pancreas Transplant Registry. Diabetologia 1991;34:S28. 17. Tamsma JT, Schaapherder A, van Bronswijk H, et al. Islet cell hormone release immediately after human pancreatic transplantation. Transplantation 1993;56:1119-23. 18. Rodriguez FJ, Toledo-Pereyra LH, Suzuki S. Twenty-fourhour total small bowel hypothermic storage preservation and transplantation in the rat: a study of various preservation solutions. J Invest Surg 1994;7:439-51.
Surgery Volume 127, Number 5 19. Balen E, Cienfuegos JA, Montuenga L, et al. Assessment of intestinal preservation injury and duodenal rejection in a multivisceral allotransplantation model in the pig. Transplant Proc 1996;28:2646-7. 20. Heise J, Sutherland DER, Heil J, et al. 72-hour preservation of pancreatic autotransplants in dogs using a urinary drainage technique. Transplant Proc 1988;20:1029. 21. Florack G, Sutherland DER, Morel P, et al. Effective preser-
Humar et al 551
vation of human pancreas grafts. Transplant Proc 1989;21:1369. 22. Florack G, Sutherland DER, Heil J, et al. Long-term preservation of segmental pancreas autografts. Surgery 1982;92:260. 23. Largiader F, Baumgartner D, Uhlschid G. Ischemia tolerance of human pancreatic transplants. Transplant Proc 1984;16:1285-6.
Background. Prolonged preservation of the donor organ may result in delayed graft function or nonfunction after most organ transplants. We studied whether or not prolonged preservation increases surgical complications after pancreas transplants. Methods. Between January 1, 1994, and September 30, 1998, a total of 294 recipients underwent cadaver pancreas transplants at our institution. Recipients were analyzed in 2 groups: those with pancreas preservation time ≤ 20 hours (n = 211) versus > 20 hours (n = 83). Results. Demographic data were similar between the 2 groups, except that mean donor age in the prolonged preservation group was significantly lower. Despite use of younger donors, prolonged preservation was associated with an increased incidence of surgical complications, most notably leaks, thrombosis, and wound infections. Grafts with prolonged preservation were more often noted by the transplant surgeon to be edematous after reperfusion, although the incidence of hyperamylasemia posttransplant did not differ between the 2 groups. Graft and patient survival rates also did not differ between the 2 groups. The incidence of early graft loss (< 3 months) was, however, higher in the prolonged preservation group (20.5% versus 9.0%, P = .04). Conclusions. Prolonged preservation of the donor organ increases the incidence of surgical complications after pancreas transplants. All attempts should be made to minimize preservation time, keeping it below 20 hours, if possible. (Surgery 2000;127:545-51) From the Department of Surgery, University of Minnesota, Minneapolis
THE INTRODUCTION OF EFFECTIVE preservation solutions, such as the University of Wisconsin solution, has allowed significant prolongation of cold ischemia for all solid organ grafts.1-3 Yet, eventually, organ damage ensues. Clinically, such damage may manifest itself with delayed graft function or nonfunction. The functional consequences of prolonged preservation have been well documented for kidney,4 liver,5 and pancreas6 transplants; but other consequences—such as surgical complications—have not been as fully described. Ischemic injury may lead to cellular swelling and graft edema,7 which may then lead to surgical complications with significant morbidity. The purpose of this study was to document the effects of prolonged preservation on the incidence of surgical complications after pancreas transplants. Attempts to shorten preservation time Accepted for publication November 25, 1999. Reprint requests: Abhi Humar, MD, Department of Surgery, University of Minnesota Medical School, Box 328, Mayo Building, 420 Delaware St S.E., Minneapolis, MN 55455. Copyright © 2000 by Mosby, Inc. 0039-6060/2000/$12.00 + 0 doi:10.1067/msy.2000.104742
11/56/104742
could be more aggressive if the impact on outcome were known to be significant. METHODS Patients. Between January 1, 1994, and September 30, 1998, a total of 294 recipients underwent cadaver pancreas transplants at the University of Minnesota. Recipients were analyzed in 2 groups: those with pancreas preservation time ≤ 20 hours (n = 211) versus > 20 hours (n = 83). Mean preservation time in the first group was 14.6 hours versus 22.8 hours in the latter group (P = .01). Transplant categories (simultaneous pancreas-kidney [SPK], pancreas after kidney [PAK], and pancreas transplant alone [PTA]) were proportionately similar between the 2 groups. Surgery. The donor and recipient operations, and selection criteria, have been detailed previously.8-10 Organs were flushed in situ with University of Wisconsin solution and preserved in cold University of Wisconsin solution on ice. For locally procured organs, the donor duodenum was flushed with an antibiotic solution before being stapled shut. In the recipient operation, the pancreas was anastomosed to the iliac vessels (on the right side if possible) after reconstruction of the SURGERY 545
546 Humar et al
donor superior mesenteric and splenic arteries with a donor iliac artery Y-graft. Exocrine secretions were managed with a duodenocystostomy or duodenoenterostomy. Intraoperative data on the appearance of the transplanted pancreas were noted by the attending surgeon and later entered into a database. Immediately after reperfusion, the consistency of the pancreas was recorded as being soft, firm, or hard. The presence or absence of edema in the pancreas, as assessed by the transplant surgeon, was also noted. Postoperative care. All recipients received induction therapy with an antilymphocyte preparation. Maintenance immunosuppression was initiated with tacrolimus, mycophenolate mofetil, and prednisone. During the first 6 months posttransplant, tacrolimus doses were adjusted to maintain serum levels of 10 to 15 ng/dL, then 8 to 10 ng/dL thereafter. Recipients received antiviral prophylaxis (acyclovir or ganciclovir) and antibacterial prophylaxis with a broad-spectrum antibiotic (imipenem, ampicillin/sulbactam, or ceftazidime). Fluconazole was also given during the first week posttransplant for antifungal prophylaxis. Most recipients also received standard prophylaxis against vascular thrombosis: low-dose heparin (300 to 500 units/hr) for the first 5 days posttransplant and acetylsalicylic acid (325 mg/d) for the first 3 months posttransplant. Recipients with a bleeding complication or any contraindication to anticoagulation did not receive low-dose heparin. Surgical complications. Surgical complications during the first 3 months posttransplant were included in this analysis.11,12 Complications included relaparotomy, vascular graft thrombosis, intraabdominal infections, bleeding, pancreatitis, and anastomotic or duodenal stump leaks. Relaparotomy was defined as any reoperative procedure involving the intraperitoneal or retroperitoneal space during the first 3 months posttransplant. Thrombosis was diagnosed by clinical symptoms, imaging study results, intraoperative findings, or a combination thereof. Thrombosis included venous and arterial thrombosis. Bleeding was defined as a significant complication if it necessitated a reoperation or a transfusion of more than 2 units of red blood cells. For all intra-abdominal infections included in this study, patients were symptomatic, with documented positive culture results. Cultures were obtained from aspirations guided by computed tomography or ultrasonography, or at the time of abdominal exploration. Anastomotic or duodenal stump leaks were diagnosed by clinical symptoms, imaging study results, laboratory findings, or a combination thereof.
Surgery May 2000
Postoperative pancreatitis was defined by a serum amylase level above 120 mg/dL for the first 5 postoperative days, accompanied by pain and tenderness over the graft. Rejection. Pancreas graft rejection after a bladder-drained transplant was suspected by a decrease in urinary amylase levels of 25% or more from baseline on 2 consecutive measurements and confirmed by biopsy in the majority of cases.13 Pancreas graft rejection after SPK transplants was frequently diagnosed by an increase in serum creatinine level; kidney biopsies were obtained whenever kidney graft rejection was clinically suspected. Treatment was generally with a course of antilymphocyte therapy. Data analysis. Categorical variables were analyzed by using the chi-square test, and when applicable, Fisher’s exact test. Continuous variables were analyzed parametrically with the Student t test. Pancreas graft and patient survival rates were calculated according to the Kaplan-Meier method. Pancreas graft loss was defined by the recipient’s return to exogenous insulin use after insulin independence. Calculation of patient survival included deaths occurring after pancreas and kidney graft loss. Survival rates were compared among groups with the generalized Wilcoxon’s test. A P value of ≤ .05 was considered significant. RESULTS Of 294 recipients, 211 received organs with a preservation time of ≤ 20 hours while 83 received organs with a preservation time of > 20 hours. Characteristics of the 2 groups are shown in Table I. The mean recipient age was the same in both groups, but the mean donor age was significantly lower in the prolonged preservation group (22.3 years versus 28.1 years, P = .04). The composition of the 2 groups by transplant category was similar, with SPK and PAK each accounting for roughly 40% of the transplants. Enteric drainage was performed in about 10% of recipients in both groups. Donor hemodynamic stability before organ procurement (as measured by the need for inotropes in the donor) was the same for both groups. Other characteristics of the donor, such as cause of death and renal function, were not significantly different between the 2 groups. Organs with lower preservation times were more likely to have been procured locally: 50.0% of organs with a preservation time of ≤ 20 hours were procured locally versus 35.4% of those with a preservation time of > 20 hours (P = .02). Early (< 3 months posttransplant) surgical complications for the 2 groups are shown in Table II.
Humar et al 547
Surgery Volume 127, Number 5 Table I. Characteristics of 2 groups Total number of recipients SPK/PAK/PTA Mean recipient age (y) Mean donor age (y) Mean preservation time (h) Enteric drainage: n (%) Donor inotrope use Cause of donor death Trauma CVA Donor creatinine (mg/dL) Local procurement
Preservation ≤ 20 h
Preservation > 20 h
P value
211 89/91/31 39.6 28.1 14.6 23 (10.9%) 56%
83 39/32/12 40.5 22.3 22.8 9 (10.8%) 58%
— NS NS .04 .01 NS NS
58% 25% 1.6 50.0%
64% 20% 1.5 35.4%
NS NS NS .02
SPK, Simultaneous pancreas-kidney; PAK, pancreas after kidney; PTA, pancreas transplant alone; CVA, cerebrovascular accident.
Table II. Surgical complications and outcomes Preservation ≤ 20 h Total number of recipients Complications Intra-abdominal infections Wound infections Bleeding Duodenal leaks Thrombosis Pancreatitis Relaparotomy rate Outcomes Acute rejection (3 m) Graft survival (3 m) Graft survival (3 y) Patient survival (3 y)
211
Preservation > 20 h 83
P value —
5.7% 10.4% 18.0% 5.2% 5.2% 11.4% 16.4%
8.4% 18.1% 18.0% 14.5% 10.8% 13.3% 19.3%
NS .07 NS .008 .08 NS .23
23.7% 91.0% 69.8% 92.1%
21.7% 79.5% 72.3% 93.5%
NS .04 NS NS
Prolonged preservation was associated with an increased likelihood for surgical complications, most notably leaks, thrombosis, and wound infections. The most dramatic difference in the 2 groups was in the incidence of duodenal leaks: 14.5% for preservation > 20 hours versus 5.2% for preservation ≤ 20 hours (P = .008). The leak rate increased proportionately with the preservation time (Fig 1). With a preservation time of < 15 hours, the leak rate was 1.4%. When the preservation time increased to 15 to 20 hours, the leak rate increased to 5.8% (P = .14); with 20 to 25 hours, 14.1% (P = .001); with 25 to 30 hours, 16.7% (P = .03). Though not as striking, a similar trend was seen when thrombosis was used as the dependent variable (Fig 2). The incidence of thrombosis was higher with preservation times of 20 to 25 hours and 25 to 30 hours, versus < 15 hours or 15 to 20 hours (P = .10). As a result of these complications, recipients in the prolonged preservation group tended to have a
higher relaparotomy rate during the first 3 months posttransplant (19.3% versus 16.4%, P = .23), but this trend was not statistically significant. The incidence of bleeding was the same in both groups. Grafts with prolonged preservation were more often noted to be edematous immediately after reperfusion, but this was a subjective evaluation by the transplant surgeon. The incidence of clinical pancreatitis, however, was not higher in the prolonged preservation group. Prolonged preservation did not have a significant impact on patient or graft (Fig 3) survival rates. At 3 years posttransplant, we noted no significant difference in either graft or patient survival rates between the 2 groups. However, the incidence of early graft loss (by 3 months posttransplant) was significantly higher with prolonged preservation. By 3 months posttransplant, 9.0% of grafts in the shorter preservation group had been lost versus 20.5% in the prolonged preservation group (P = .04). Most of
548 Humar et al
Surgery May 2000
Fig 1. Incidence of leaks versus preservation time. Total number of recipients in each group is shown. Asterisk, P < .05 versus incidence of leaks in < 15 hours preservation group.
Fig 2. Incidence of thrombosis versus preservation time. Total number of recipients in each group is shown.
these early graft losses represent failures secondary to technical complications, such as thrombosis or leaks, that lead to graft removal. DISCUSSION Pancreas transplants are still associated with the highest surgical complication rate of all routinely performed solid organ transplants.14 Several risk
factors have been identified that predict an increased likelihood for surgical complications, including older donor age, older recipient age, enteric (versus bladder) drainage, graft pancreatitis, and retransplants.11,12,14 Prolonged preservation and poor pancreas function have been associated,6,15 but the impact on surgical complications has not been fully described. This impact is impor-
Surgery Volume 127, Number 5
Humar et al 549
Fig 3. Pancreas graft survival rates.
tant to document: If it is significant, more aggressive efforts could be made to decrease preservation time. Previous reports have shown some influence of prolonged preservation on surgical complications such as thrombosis.16 In an earlier study from our center, logistic regression analysis revealed prolonged preservation to be a significant risk factor for death in PTA recipients.14 Our current study also suggests that prolonged preservation is a risk factor for surgical complications after pancreas transplants. Our 2 groups were demographically similar. Their composition by transplant category was nearly identical, with SPK and PAK each accounting for roughly 40% of the transplants in both groups. This fact is important to note, because the surgical complication rate is generally higher after a double-organ transplant than after single-organ transplants (PAK or PTA). The methods of exocrine drainage may also influence the complication rate. Enteric drainage and bladder drainage have slightly different complication rates. We have recently started to use enteric drainage more frequently, especially with SPK transplants where the serum creatinine level can be used to monitor for rejection. In this study, the percentage of recipients with enteric drainage was nearly identical for both groups. Donor factors may also influence surgical complications. Unstable donors may have organs with more ischemic damage. Using inotrope administration and mean serum creatinine in the donor as a crude approximation of donor stability, we were
not able to identify a significant difference in our 2 groups. Pancreases with prolonged preservation more commonly were imported from an outside region. This factor may have some impact: experience of the procuring team may have an impact on the surgical team posttransplant. Organs damaged at the time of procurement may lead to increased complications after the transplant. It is possible that the higher complication rate seen in the prolonged preservation group may be in part related to the higher percentage of imported organs in that group. However, as the number of centers performing pancreas transplants has dramatically increased, so has the number of surgeons capable of safely procuring a pancreas for transplant. Another difference between the 2 groups was that donors in the prolonged preservation group were significantly younger than in the other group. Therefore, a longer preservation time was accepted with younger donors. Earlier analyses from our center have consistently demonstrated that older donor age is a significant risk factor for surgical complications after pancreas transplants.11,12 But, even with younger donors, our current study shows that prolonged preservation was associated with an increased risk for surgical complications. Leaks, thrombosis, and wound infections were overall more common in the prolonged preservation group. These surgical complications, all associated with significant morbidity, accounted for the higher reoperation rate in the prolonged preservation group. For certain complications, such as leaks, there was very good correlation between
550 Humar et al
increasing preservation times and increasing incidence of the complication (Fig 1). This illustrates the importance of trying to minimize preservation times, rather than simply targeting for a maximum allowable time limit. The reasons for a higher surgical complication rate after prolonged preservation are unclear, but may in part be related to increased graft edema after prolonged preservation. Reperfusion after cold preservation results in some tissue injury, which manifests as graft edema and swelling.7,17 The significantly higher leak rate with prolonged preservation may also be attributable to a different sensitivity of the donor duodenum (versus the pancreatic parenchyma) to prolonged preservation. Data from the small bowel literature suggest that the bowel has a poor tolerance to prolonged preservation.18,19 Since the vast majority of pancreas transplants are performed with the donor duodenum intact, this may explain the higher incidence of leaks with increasing preservation times. We did not examine the effect of prolonged preservation on graft endocrine function, because our purpose was to examine its impact on surgical complications. In previous analyses from our center, preservation for more than 24 hours was possible, but did result in a trend toward slightly inferior endocrine function; a higher number of such recipients had abnormal oral glucose tolerance tests postoperatively.6,20-22 Other studies have also demonstrated decreased insulin secretion from organs preserved for more than 24 hours.15,23 We were not able to demonstrate a significant detrimental effect of prolonged preservation on long-term (ie, 3-year) graft survival. However, the 3month graft survival rate in the prolonged preservation group was significantly lower. These early graft losses reflect the higher surgical complication rate in this group and the consequence of adverse events such as thrombosis or leaks. In conclusion, preservation time has an impact on the development of surgical complications after pancreas transplants. Prolonged preservation, likely through the development of graft edema, leads to an increased incidence of complications, such as thrombosis and leaks. The incidence of surgical complications after pancreas transplants remains high, so all efforts must be made to minimize preservation time (preferably to < 20 hours). Special thanks to Mary Knatterud and Ernestine Draper for their invaluable help in the preparation of this manuscript.
Surgery May 2000 REFERENCES 1. Belzer FO, Southard JH. Principles of solid-organ preservation by cold storage. Transplantation 1988;45:673-6. 2. D’Alessandro AM, Stratta RJ, Sollinger HW, et al. Use of UW solution in pancreas transplantation. Diabetes 1989;38:7-9. 3. Abouna GM, Sutherland DER, Florack G, et al. Function of transplanted human pancreatic allografts after preservation in cold storage for 6 to 26 hours. Transplantation 1987;43:630-6. 4. Koning OH, Ploeg RJ, van Bockel JH, et al. Risk factors for delayed graft function in cadaveric kidney transplantation: a prospective study of renal function and graft survival after preservation with University of Wisconsin solution in multiorgan donors. Transplantation 1997,63:1620-8. 5. Porte RJ, Ploey RJ, Hansen B, et al. Long-term graft survival after liver transplantation in the UW era: late effects of cold ischemia and primary dysfunction. Transplant Int 1998,11:S164-7. 6. Morel P, Moudry-Munns K, Najarian JS, et al. Influence of preservation time on outcome and metabolic function of bladder-drained pancreas transplants. Transplantation 1990,49:294-303. 7. Schaapherder AF, de Roos A, Shaw PC, et al. The role of early baseline computed tomography in the interpretation of morphological changes after kidney-pancreas transplantation. Transplant Int 1993,6:270-6. 8. Gruessner RWG, Sutherland DER. Pancreas transplantation: Part I: The donor operation. Surg Rounds 1994;17:311-24. 9. Gruessner RWG, Sutherland DER. Pancreas transplantation: Part II: The recipient operation. Surg Rounds 1994; 17:383-91. 10. Gruessner RWG, Dunn DL, Gruessner AC, et al. Recipient risk factors have an impact on technical failure and patient and graft survival rates in bladder-drained pancreas transplants. Transplantation 1994;57:1598-1606. 11. Troppmann C, Gruessner AC, Sutherland DER, et al. Surgical complications requiring early relaparotomy after pancreas transplantation: a multivariate risk factor and economic impact analysis. Ann Surg 1998;227:255-68. 12. Troppmann C, Gruessner AC, Benedetti E, et al. Vascular graft thrombosis after pancreatic transplantation: univariate and multivariate operative and nonoperative risk factor analysis. J Am Coll Surg 1996;182:285-316. 13. Gruessner RWG, Sutherland DER. Clinical diagnosis in pancreatic allograft rejection. In: Solez K, Racusen LC, Billingham ME, editors. Solid organ transplant rejection: mechanism, pathology, and diagnosis. New York: Marcel Dekker, Inc; 1996. p. 455-89. 14. Gruessner RWG, Sutherland ER, Troppmann C, et al. The surgical risk of pancreas transplantation in the cyclosporine era: an overview. J Am Coll Surg 1997;185:128-43. 15. Lopez-Medrano RM, Ponce J, Varela G, et al. Extensive decrease in insulin secretion by the pancreas preserved for 24 hours in UW-1 solution. Transplant Proc 1991;23:1676-8. 16. Sutherland DE. Report from the International Pancreas Transplant Registry. Diabetologia 1991;34:S28. 17. Tamsma JT, Schaapherder A, van Bronswijk H, et al. Islet cell hormone release immediately after human pancreatic transplantation. Transplantation 1993;56:1119-23. 18. Rodriguez FJ, Toledo-Pereyra LH, Suzuki S. Twenty-fourhour total small bowel hypothermic storage preservation and transplantation in the rat: a study of various preservation solutions. J Invest Surg 1994;7:439-51.
Surgery Volume 127, Number 5 19. Balen E, Cienfuegos JA, Montuenga L, et al. Assessment of intestinal preservation injury and duodenal rejection in a multivisceral allotransplantation model in the pig. Transplant Proc 1996;28:2646-7. 20. Heise J, Sutherland DER, Heil J, et al. 72-hour preservation of pancreatic autotransplants in dogs using a urinary drainage technique. Transplant Proc 1988;20:1029. 21. Florack G, Sutherland DER, Morel P, et al. Effective preser-
Humar et al 551
vation of human pancreas grafts. Transplant Proc 1989;21:1369. 22. Florack G, Sutherland DER, Heil J, et al. Long-term preservation of segmental pancreas autografts. Surgery 1982;92:260. 23. Largiader F, Baumgartner D, Uhlschid G. Ischemia tolerance of human pancreatic transplants. Transplant Proc 1984;16:1285-6.