surgical complications in solitary pancreas and combined pancreas-kidney transplantations

Surgical Complications in Solitary Pancreas and Combined Pancreas-Kidney Transplantations

Claire F. Ozaki, Mt~, Rolaert J. titratta, MD, Rodney J. Taylor, Mr), Alan N. LangnaS, DO, J. Steven Bynon, MD, Byers W. Shaw, Jr., MD, Omaha, Nebraska

The benefits of pancreas transplantation (PT) must be weighed against the morbidity associated with the operative procedure and long-term immunosuppression. Over a 32-month period, we performed 73 PTs including 61 combined pancreas- k idn~ transplants (PKT) and 12 solitary PTs. In the PKT group, 25 reoperations were performed in 18 ~atients (29.5%) at a mean of 39 4. 12 days after transplant. In the solitary PT group, 16 reoperations were performed in 8 recipients (66.7%, p = 0 .03) at a mean of 87 4- 12 days after PT (p < 0 , 0 i ) . In the PKT group, pancreas allograft survival was 93.4%. Vascular thrombosis resulted in the loss of two pancreas allografts. In the solitary PT group, pancreas allograft survival was 50% (p <0 .001 ), with 6 transplant pancreatectomies performed for either infections (5) or vascular (1) complications. Surgical complications after PT are common (35.6% in this series), occur earlier in patients who undergo PKT, and are more frequent and morbid in patients undergoing solitary PT, especially after a previous kidney transplant. An aggressive surgical approach can lead to a high rate of pancreas allograft salvage without jeopardizing either the patient or the renal allograft.

From the Department of Surgery, University of Nebraska Medical Center, Omaha, Nebraska.

Requests for reprints should be addressed to Robert J. Stratta, MD, Department of Surgery, University of Nebraska Medical Center, 600 South 42nd Street, Omaha, Nebraska 68198-3280.

Presented at the 44th Annual Meeting of the Southwestern Surgi- cal Congress, Sc.ottsdale, Arizona, April 26-29, 1992.

W ith improvements in organ retrieval technology, refinements in surgical techniques, and advances

m clinical immunosuppression, allograft survival rates have improved dramatically, which has resulted in an increased interest in vascularized pancreas transplantation (PT).As of September 30, 1991, 3,621 PTs had been performed worldwide [1]. I n 1991, 537 PTs were performed among about 60 different transplant centers in the United States [2]. Approximately 80% of PTs have been performed in conjunction with a simultaneous kidney transplant [combined pancreas-kidney transplant (PKT)], whereas the remaining 20% have been divided between nonuremic diabetic patients who received a pancreas transplant alone and diabetic patients With a functioning kidney transplant who then received a solitary pancreas-after-kidney (PAK) transplant [3].

Since immunosuppression is necessary to prevent allograft rejection, widespread application of PT to the type I diabetic population is not appropriate at this time. Anoth- er major problem of PT is the morbidity associated with the operative procedure. However, as PT has evolved, technical problems have been overcome, and the procedure has emerged as an effective therapeutic option i n carefully selected insulin-dependent type I diabetic patients: The purpose of this study is to: (1) report the incidence of surgical complications after PT; (2) determine the etiology and significance of complications war- ranting reoperation; and (3) identify potential risk factors that might place patients at an increased risk for these complications.

PATIENTS AND METHODS Study population and design: From April 1989

through December 1991, a total of 73 consecutive vascularized PTs were performed at our center. During this period, data from all recipients were retrospectively ana- lyzed for this study. The minimum length of follow-up was 3 months.

Demographic and clinical characteristics of the recipient population are summarized in Table i. For Study purposes, the population was divided into the 61 patients undergoing PKT and the 12 patients receiving a solitary PT (6 PAK, 6 PT alone), The age range was 24 to 54 years, and the severity of type I diabetes was assessed by the mean duration, daily insulin dose, and glycosylated hemoglobin levels. All patients had clinically significant retinopathy, and the majority of patients had significant peripheral and/or autonomic neuropathy. A total of 32 concomitant procedures were performed in 27 patients at the time of transplant, including peritoneal dialysis catheter removal (19), transplant and/or native nephrectomy (6), cholecystectomy (3), tubal ligation (2), appendecto- my (1), and tracheostomy (1).

546 THE AMERICAN JOURNAL OF SURGERY VOLUME 164 NOVEMBER 1992

COMPLICATIONS OF PANCREAS AND KIDNEY TRANSPLANTATION

Organ procurement and preservation: In aU cases, combined hepatectomy and pancreaticoduodenosplenec- tomy was performed via an en bloc technique in conjunction with multiple organ retrieval [4]. Bilateral nephrectomy was also performed in an en bloc fashion. The organs were then separated on the back table, and the liver, pancreas, and kidneys were packaged separately and sterilely under cold storage conditions in University of Wisconsin (UW) solution (Viaspan, DuPont Pharma- ceuticals, Waukegan, IL). Cold ischemia times are shown in Table I.

Recipient selection, preparation, and operation: Patients were selected for transplantation based on ABO blood type compatibility, length of time on the waiting list, a negative T-lymphocytotoxic crossmatch, medical urgency, and HLA-DR matching. HLA matching data are depicted in Table I. All patients underwent preopera- tive selective bowel decontamination with a mechanical bowel preparation (Go-Lytely, Braintree Laboratories, Braintree, MA) and an antibiotic bowel preparation (oral clindamycin and neomycin). Prior to the recipient operation, the kidney and pancreas were prepared for transplantation on the back table under cold storage conditions [5]. The arterial circulation to the pancreas is recon- structed with a donor iliac artery bifurcation "Y" auto- graft to the splenic and superior mesenteric arteries. The ends of the duodenum are mobilized, stapled, and then oversewn, leaving the second portion as a duodenal segment attached to the pancreas [6]. The splenic hilum is then partially mobilized from the tail of the pancreas, but the spleen is left attached so that it can be used as a handle.

After preparation of the organs, the recipient operation is performed through a midline approach [7]. The bladder is mobilized from its lateral attachments, and the bowel is mobilized and retracted into the upper abdomen. The pelvic peritoneum overlying the iliac vessels is in- cised, and the external iliac artery and vein are mobilized. In PKT cases, the kidney transplant is performed first to the left iliac vessels with cod-to-side vascular anastomoses and an extra-vesical ureteroneocystostomy. The right iliac artery is then completely mobilized with preservation of the hypogastric artery. This enables complete mobilization of the right iliac vein with ligation and divi- sion of the hypogastric vein and any other posterior branches. The pancreas is then transplanted to the right lilac vessels with end-to-side vascular anastomoses including a tension-free venous reconstruction with the iliac vein positioned lateral to the artery (Figure 1). In patients undergoing solitary PT, the iliac vessels on the side of the transplant are completely mobilized as previously noted. A two-layer duodenocystostomy is then performed with a running inner layer of 3-0 Maxon (Davis and Geck, Wayne, N J) and an outer layer of interrupted 3-0 PDS (Ethicon, Inc., Somerville, N J) [8]. The pancreas is placed in the paracolic gutter, and hemostasis is achieved. Silastic active sump drains are placed medial and lateral to the pancreas and medial to the kidney (if transplanted) and exit through separate stab incisions in the lateral abdominal wall as the wound is dosed.

TABLE l Demographic and Clinical Characteristics

Pancreas-Kidney Solitary Pancreas Transplant Transplant

No. of patients 61 12 Age (y) (mean)(range) 35.8 (24-51) 39.0 (26-54) Gender: Male (%) 36 (59) 6 (50)

Female (%) 25 (41) 6 (50) Duration of diabetes (y) 23.6 (13-41) 26.8 (14-47)

(mean) (range) Insulin dose (U/d) 41 (12-100) 41 (15-75)

(mean) (range) Glycosylated hemoglobin 9.9 -+ 0.3 10,8 _+ 0.8

level (%) Prior dialysis or kidney 32 (52.5) 6 (50)

transplant Retinopathy (%) 61 (100) 12 (100) Mean cold ischemia time (h)

Pancreas 16.3 --- 0.5 17.4_ + 1.0 Kidney 15.7 _+ 0.5 - -

HLA-match (mean) (range) AB 0.79 (0-2) 1.3 (1-2) DR 0.52 (0-2) 1.2 (1-2)

Operating time (h) 5 (3.5-7) 3 (2-4) (mean) (range)

Operative transfusions (U) 2 (0-6) 1 (0-2) (mean) (range)

All patients were managed with a central line and received intraoperative volume loading with colloid and liberal use of mannitol and furosemide to promote a di- uresis and minimize pancreatic edema. Most patients also received low-dose dopamine (3 to 5 gg/kg/min) in an attempt to optimize renal function. Prior to revascu- larization, most patients were managed with a continuous intravenous insulin infusion. Operating time and transfu- sion data are shown in Table I.

Perioperative management and immunosuppression: Perioperative antibiotics consisted of a single preop- erative, intraoperative, and. three postoperative doses of cefazolin (1 g intravenously). Patients also received a single postoperative dose of vancomycin (1 g intravenously) and daily doses of ketoconazole (400 mg orally) or fluconazole (200 mg orally) for 3 days. All patients also received oral nystatin, acyclovir, and trimethoprim/sul- famethoxazole prophylaxis. In PKT patients, anticoagu- lant therapy was not used, but all patients did receive antiplatelet therapy consisting of oral aspirin (80 mg/d) and dipyridamole (50 mg 3 times a day). In solitary PT cases, patients were systemically administered heparin, with 5,000 U of intravenous heparin intraoperatively, which was not reversed. Low-dose continuous infusion heparin (500 U/h intravenous) was then continued for 5 to 7 days. In both groups, nasogastric tube decompression was maintained for 2 to 3 days and Foley catheter drainage for 7 to 10 days. Intra-abdominal drains were usually removed within 5 days.

All patients received quadruple immunosuppression with OKT3 induction in combination with cyclosporine, steroids, and azathioprine [9,10]. Initial rejection epi- sodes were treated with a steroid pulse and/or recycle.

THE AMERICAN JOURNAL OF SURGERY VOLUME 164 NOVEMBER 1992 547

OZAKIET AL

DOU~dO~e n ocystostom/..j~ ~ ~~~, ~ !!! ~ Figure 1, Surgical technique of combined pancreas-kidney transplant through a midline Intraperitoneel ap- preach. A solitary pancreas transplant was performed in a similar fashion except for omission of the kidney transplant. See text for details.

TABLE II Surgical Complications

Solitary PT PT

PKT Alone PAK (Total) (n = 61) (n = 6) (n = 6) (n = 12)

No. of compli- 25 (18 pts) 3 (2 pts) 13 (6 pts) 16 (8 pts) cations (29.5%) (33.3%) (100%)* (66.7%)

Vascular complications Total 8 (13.1%) 0 3 (50%)* 3 (25%) Hemorrhage 4 0 1 1 Thrombosis 2 0 1 1 Arteriovenous 2 0 0 0

fistula Pseudo- 0 0 1 1

aneurysm

Technical complications Total 8 (7 pts) 1 (16.7%) 0 1 (8.3%)

(11.5%) Duodenal 4 1 0 1

segment leak Ureter leak 4

Infectious complications With pancre- 3 (4.9%) 1 (16.7%) 7 (5 pts) 8 (5 ~ots)

atitis (83.3%)* (41.7%)

Other complications 6 (9.8%) 1 (16.7%) 3 (50%)* 4 (33.3%)

PKr = pancreas-kidney transplant; PT = pancreas transplant: PAK = pancreas- after-kidney transplant; pts = patients

*p <0.05 (PAKversus PKT).

Minnesota ALG or Atgam (Upjohn Co., Kalamazoo, MI) was employed for the treatment of steroid-resistant rejection or for rejection occurring during OKT3 induction therapy.

After transplantation, duplex ultrasonography and radionuclide perfusion scans of the pancreas (and kidney when transplanted) were performed on the first postoperative day and whenever clinically indicated. A surgical complication was defined as any and all instances in which the patient was returned to the operating room for reoperation.

Statistical analysis: Data are reported as mean 4- SEM. Univariate analysis was performed by the Stu- dent's t-test for continuous variables and Fisher's exact test for categorical variables. Patient and graft survival are presented as actuarial results as computed by Wil- coxon life-table analysis. A p value less than 0.05 was considered significant.

RESULTS Patient and graft survival: In the PKT group, pa-

tient survival was 98%, kidney allograft survival was 97%, and pancreas allograft survival was 93% after a mean follow-up of 15 months (range: 3 to 35 months). Four pancreas and two kidney allografts were lost. Two patients developed venous thrombosis of the pancreas allograft on the first postoperative day, resulting in allograft pancreatectomy. One patient developed refractory pancreas rejection at 7 months, whereas another patient developed refractory kidney rejection at 2 months. One patient died at home of a myocardial infarction 1 month after transplantation with functioning pancreas and kidney allografts.

In the solitary PT group, patient survival was 83%, and pancreas allograft survival was 50% (p <0.001 versus pancreas survival after PKT) after a mean follow-up of 8 months (range: 3 to 12 months). Six pancreas allografts were removed for infection (five) or arterial thrombosis (one~. One patient died of a massive pulmonary embolism 5 months after transplantation and 1 week after transplant pancreatectomy for sepsis, Another patient died of uncontrolled sepsis 3 months after transplant pancreatectomy for sepsis. Pancreas allograft survival was significantly better in recipients of PT alone (5 of 6, or 83%) than patients who underwent PAK transplantation (1 of 6, or 17%, p ffi 0.04).

Operative complications after PKT: In the PKT group, 25 reoperations were performed in 18 patients (30%) at an average of 39 4- 12 days after transplant. Indications for reoperation are listed in Table IL A total of eight reoperations were performed for vascular complications (four for hemorrhage, two for thrombosis, two for



arteriovenous fistulas) at a mean of 15 days after PKT. Eight reoperations were performed for technical problems (ureter or duodenal segment leaks) at a mean of 20 days after transplant. Three patients (5%) had reoperations for intra-abdominal and wound infections in associ- ation with pancreatitis. The remaining six reoperations included: removal of a retained common duct stent, drainage of an obstructing pelvic lymphocele, enterolysis for small bowel obstruction, orchiectomy for testicular abscess, and enteric conversion of the exocrine pancreas for persistent bematuria due to ulceration with subsequent reconversion to bladder drainage after development of an enteric anastomotic leak. No patients or renal allografts were lost due to operative complications. Although two pancreas allografts were removed because of vascular thrombosis, none were lost as a result of either infectious or local complications.

Operative complications after solitary PT: A total of 16 reoperations were performed in 8 patients (67%) after solitary PT. The incidence of operative complications was significantly higher in recipients of PAK transplants versus either PT alone or combined PKT (Table II). Indications for reoperation are listed in Table II. Operative complications tended to occur much later after solitary PT (mean: 87 4- 12 days versus 39 days after PKT, p <0.01). Despite their later occurrence, reoperations after solitary PT were associated with much greater morbidity including two deaths and six transplant pancreatectomies. The incidence of vascular complications was higher in PAK recipients (50% versus 13% PKT, p = 0.05) and usually occurred early (mean: 26 days). Com- pared with PKT, technical morbidity (leaks) had a similar incidence after solitary PT. However, infectious complications were a major problem since 8 of the 16 reoperations (including 5 transplant pancrcatectomies) were performed because of infection with associated pancreatitis. The incidence of major intra-abdominal or wound infections requiring operative intervention was much higher in PAK recipients (Table II). The remaining operations performed after solitary PT included cholecystectomy for acalculous cholecystitis (two) and enterolysis for small bowel obstruction (two). Miscellaneous procedures were more common in the PAK group (50% versus 10% PKT, p = 0.03).

Overall results- For the entire group (73), overall pancreas allograft survival was 86% after a mean follow- up of 14 months. Ten pancreas allografts were lost (1 death, 1 rejection, 3 thromboses, and 5 peripancreatic infections with pancreatitis). In patients with vascular complications, pancreas allograft survival was 50%. In patients with intra-abdominal infections, pancreas alto- graft survival was 40%. Transplant pancreatectomy was performed much earlier for vascular versus infectious complications (mean: 5 days versus 108 days, p <0.01). In comparing the 47 patients without reoperations to the 26 patients (36%) with surgical complications, no major risk factors were identified other than type of transplant.

The incidence of rejection and the overall level of immunosuppression were similar in patients with and without operative complications. However, the develop-

ment of allograft pancreatitis, especially with infection, usually resulted in surgical morbidity.

COMMENTS Until recently, the results of vascularized PT were

inferior to the excellent results achieved in kidney, heart, and liver transplantation. However, a number of recent innovations have enhanced the safety and reliability of PT including: (1) the introduction of bladder drainage of the exocrine secretions [8,11]; (2) the evolution from segmental to whole-organ pancreas to pancreaticoduo- denal transplantation [3]; (3) the conversion from a duodenal button to a duodenal segment technique of bladder drainage [6,8]; (4) the application of quadruple immuno- suppressive therapy [9,10]; (5) the success of combined PKT [3,7,12-18]; and (6) the introduction of UW solution for extended preservation of the pancreas [5,19]. Despite these advances, the surgical complication rate has remained unacceptably high and has precluded widespread application of this therapy to type I diabetic patients.

In contrast to other solid organs transplanted, the pancreas is susceptible to a unique set of complications because of its exocrine elements and low blood flow. Allo- graft pancreatitis has become less of a problem with: (1) the introduction of UW preservation solution [5,19]; (2) the "no-touch" technique of donor duodenopancreatico- splenectomy using the spleen and duodenum as handles [4,5]; (3) intraperitoneal, whole-organ transplantation with active drainage [18,20]; and (4) selective use of a long-acting somatostatin analogue (Sandostatin, Sandoz Pharmaceuticals, Raritan, N J) [21]. However, complications related to the exocrine pancreas remain a major cause of morbidity after PT [6,8,18,21,22]. Fortunately, most cases of allograft pancreatitis are self-limited and due either to early (ischemic) injury, rejection, or reflux pancreatitis. Surgical problems related to exoerine pancreatic drainage and allograft pancreatitis are usually due to technical problems (leak, fistula) or infection with development of fluid collections, pseudoeysts, or abscesses surrounding the pancreatic graft [23]. In these instances, operative exploration is often required, especially in the setting of infection. In our experience, the presence of peripancreatic infection portends a poor prognosis in over 50% of cases, resulting in multiple procedures and eventual allograft pancreatectomy.

Pancreas allografts are prone not only to pancreatitis but also to other vascular problems because of the intrin- sically low microcirculatory flow of the organ. However, the incidence of vascular complications has decreased because of better preservation, less pancreatitis, the rou- tine use of an arterial graft, a tension-free portal venous anastomosis after complete mobilization of the recipient iliac vein, and antiplatelet (but not full anticoagulation) therapy to reduce the risk of thrombosis without increas- ing the risk of hemorrhage. Whether vascular thrombosis occurs as a manifestation of preservation or immunologic injury or is due to a technical problem is difficult to determine [24,25]. The two eases of venous thrombosis after PKT were probably technical in origin, whereas the

THE AMERICAN JOURNAL OF SURGERY VOLUME 164 NOVEMBER 1992 5 4 9

OZAKI El" AL

one case of arterial thrombosis after solitary PT may have been related to other factors. The reoperations for hemorrhage were usually due to bleeding from the mesenteric bundle, which was also the site of the two arteriovenous fistulas. These complications can be minimized by careful hemostasis, tension-free vascular anastomoses, and indi- vidual ligation of the superior mesenteric artery and vein.

Other investigators have reported a high incidence of wound or infectious complications after PT that are not necessari ly related to immunosuppression [7,12-18,20,26]. Important factors in preventing wound or intraperitoneal infections include: (1) intraperitoneal placement of the organ(s) through a midline approach [20]; (2) active drainage of the pelvis to prevent fluid collections; (3) selective use of Sandostatin [2I]; and (4) selective bowel decontamination with appropriate antibiotic coverage. Exocrine leakage remains a problem, although we had no cases of anastomotic leakage. All duodenal segment leaks were related either to apparent devascularization injuries during the back table dissec- tion (four) or cytomegalovirus ulceration (one). The four ureter leaks were due to excessive tension possibly caused by manipulation of the bladder for PT. These problems can be reduced by meticulous surgical technique and the use of slowly absorbable monofilament sutures.

It was somewhat surprising that PKT recipients actu- ally fared better not only from a patient- and graft-survival standpoint but also in terms of fewer surgical complications. Since combined PKT is a longer and more technically extensive operation, we anticipated that it would be attended by a higher reoperation rate, especially in the setting of uremia or concurrent dialysis. However, a striking finding of this study was the high incidence of surgical complications that occurred in recipients of a PAK transplant. These six patients had previously under- gone cadaveric kidney transplantation 4 to 188 months prior to solitary PT. The demographic and clinical characteristics of this group were similar to those of the other groups except for the presence of long-term immunosuppression. The only other distinguishing feature between solitary PT and PKT recipients was the use of short-term anticoagulation with heparin in the solitary PT group. Whether this may have predisposed these patients to the development of hemorrhagic pancreatitis remains specu- lative.

In addition to pancreatitis, the other major problem that occurred in PAK recipients was the subsequent development of peripancreatic infections. Many of these patients had a prior history of major infections, which we currently view as a contraindication to PAK transplantation. Other series have reported more favorable results in PAK recipients versus PT alone recipients, since these patients are already committed to long-term immunosuppression [3,12,27]. We suspect that our poor results in this subgroup are predominantly due to recipient selection.

In summary, surgical complications after PT are common (36% in this series), occur earlier in PKT recipients, and are more frequent and morbid in patients receiving solitary PT, especially after a previous kidney transplant.

Early diagnosis is critical to graft salvage, but prompt surgical intervention may still result in a high rate of graft loss when vascular or infectious complications are pre- sent. However, an aggressive surgical approach can lead to a high rate of pancreas allograft salvage in most other cases without jeopardizing either the patient or the renal anograft.

REFERENCES 1. Moudry-Munns KC, Sutherland DER. International pancreas transplant registry newsletter 1991; 4(1): Appendix 2. 2. United Network for Organ Sharing Update 1992; 8: 5. 3. Suthcrland DER, Gillingham K, Moudry-Munns KC. Results of pancreas transplantation in the United States for 1987-90 from the United Network for Organ Sharing (UNOS) Registry with comparison to 1984-87 results. Olin Transplantation 1991; 5: 330-341. 4. Stratta R J, Taylor R J, Specs EC, et al. Refinements in cadaveric pancreas-kidney procurement and preservation. Transplant Proc 1991; 23: 2320-22. 5. Sollinger HW, Vernon WB, D'Alessandro AM, et al. Combined liver and pancreas procurement with Belzcr-UW solution. Surgery 1989; 106: 685-91. 6. D'Alessandro AM, Sollinger HW, Stratta ILl, et al. Comparison between duodenal button and duodenal segment in pancreas transplantation. Transplantation 1989; 47: 120-2. 7. Sollinger HW, Stratta R J, D'Alessandro AM, et al. Experience with simultaneous pancreas-kidney transplantation. Ann Surg 1988; 208: 475-83. 8. Sollinger HW, Pirsch JD, D'Alessandro AM, et al. Advantages of bladder drainage in pancreas transplantation: a personal view. Clin Transplantation 1990; 4: 32-6. 9. Sollinger HW, Stratta R J, Kalayogh M, et al. Pancreas transplantation with pancreatieocystostomy and quadruple immunosuppression. Surgery 1987; 102: 674-9. 10. Melzer JS, D'Alessandro AM, Kalayoglu M, et al. The use of OKT3 in combined pancreas-kidney allotransplantation. Trans- plant Proc 1990; 22: 634-5. 11. Sollinger HW, Cook K, Kamps D, et al. Clinical and experi- mental experience with pancreatieo-eystostomy for exocrine pancreatic drainage in pancreas transplantation. Transplant Proc 1984; 16: 749. 12. Sutherland DER, Dunn DL, Goetz FC, et al. A ten-year experience with 290 pancreas transplants at a single institution. Ann Surg 1989; 210: 274-88. 13. Corry R J, Nghiem DD, Schulak JA, et al. Surgical treatment of diabetic nephropathy with simultaneous pancreatieoduodenal and renal transplantation. Surg Gynecol Obstet 1986; 162: 547-55. 14. Cosimi AB, Auchincloss H, Delmonico FL, et al. Combined kidney and pancreas transplantation in diabetics. Arch Surg 1988; 123: 621-5. 15. Garvin PJ, Carney KM, Aridge D, Evolution of synchronous renal and pancreatic transplantation. Am J Surg 1989; 158: 625-9. 16. Schulak JA, Mayes JT, Hricik DE. Combined kidney and pancreas transplantation: a safe and effective treatment for diabetic nephropathy. Arch Surg 1990; 125: 881-5. 17. Stratta R J, Taylor R J, Zorn BH, et al. Combined pancreas- kidney transplantation: preliminary results and metabolic effects. Am J Gastroenterol 1991; 86: 697-703. 18. Sollinger HW, Kneehtle S J, Reed A, et al. Experience with 100 consecutive simultaneous kidney-pancreas transplants with bladder drainage. Ann Surg 1991; 214: 703-1t. 19. D'Alessandro AM, Stratta RJ, Sollinger HW, et al. Use of UW solution in pancreas transplantation. Diabetes 1989; 38 (Suppl 1): 7-9. 20. Tesi R J, Henry ML, Elkhammas EA, Sommer B J, Ferguson



RM. Decreased wound complications of combined kidney/pancreas transplant using intraabdominal pancreas graft placement. Clin Transplantation 1990; 4: 287-91. 21. Stratta R J, Taylor RJ. The role of Sandostatin rescue in vascularized pancreas transplantation. Transplant Proc 1993 (In press). 22. Stratta R J, Sollinger HW, Groshek M, et al. Differential diagnosis of hyperamylasemia in pancreas allograft recipients. Transplant Prec 1990; 22: 675-7. 23. Hesse U J, Sutherland DER, Najarian JS, Simmons RL. Intra- abdominal infection in pancreas transplant recipients. Ann Surg 1986; 203: 153-62. 24. Manner M, Schult W, Senninger N, et al. Evaluation of preservation damage after porcine liver transplantation by assessment of hepatic microcirculation. Transplantation 1990; 50: 940-3. 25. Mazzaferro V, Esquivel CO, Makowka L, et al. Hepatic artery thrombosis after pediatric liver transplantation: medical or surgical event? Transplantation 1989; 47: 971-7. 26. Barker R J, Mayes JT, Schulak JA. Wound abscesses following retroperitoneal pancreas transplantation. Clin Transplantation 1991; 5: 403-7. 27. Sutherland DER, Kendall DM, Moudry KC, et al. Pancreas transplantation in nonuremic type I diabetic recipients. Surgery 1988; 104: 453-64.

DISCUSSION Paul J. Garvin (St. Louis, MO): In your manuscript,

you state that several modifications have contributed to the current success of pancreas transplantation, including intraperitoneal placement of the allograft, the use of the duodenal segment, the introduction of University of Wis- consin solution, and many other advances. Although I concur that these modifications certainly have contributed to our current success with this procedure, I would like to issue a note of caution. The majority of these modifications have been identified by trial and error. With the large number of transplants being performed nationally and internationally, these modifications should be studied in a prospective fashion. I have several ques- tions. First, since you state that vascular thrombosis seems to be primarily a technical problem related to how the venous and arterial anastomoses are placed, do you believe that anticoagulation of any kind, even with antiplatelet agents, has a role in pancreas transplantation, particularly since several of the complications were related to hemorrhage, and anticoagulation may have contributed to this problem. Secondly, do you think that the more vigorous anticoagulation with heparin in the solitary pancreas group contributed to the higher incidence of infectious complications, perhaps by resulting in small hematomas around the allograft? In addition, no mention is made of the group of patients without surgical complications. It would seem helpful to analyze certain variables, like the incidence of acute rejection, cumulative immunosuppression, and graft pancreatitis, or other vari-

ables that may have accounted for why this group did not develop surgical complications. Lastly, based on your ret- rospective experience, do you now believe that pharmaco- logic exocrine suppression with octreotide acetate, or some other agent, may have a role during the early post- transplant period to reduce the incidence of surgical complications that must be related, at least in part, to the exocrine pancreas.

Jon S. Thompson (Omaha, NE): I noticed that there were several complications related to the duodenal seg- ments. Were any of these related to rejection of the intes- tinal segment, and, if so, was the incidence different in the patients with the solitary pancreas transplants and those with the combined transplants?

John J. Brems (San Diego, CA): As pancreas transplantation becomes more accepted, one of the things that will probably determine its continued acceptance will be whether the end-organ damage resulting from diabetes, such as retinopathy or neuropathy, can be halted or reversed. I wonder if you have information on long-term follow-up of your patients, as related to the reversal of retinopathy or neuropathy.

Claire Ozaki (closing): With respect to the incidence of vascular thromboses and our prophylaxis, all patients are given antiplatelet therapy, including aspirin and dipyridamole, immediately after transplant. We have been administering heparin to patients in the pancreas transplant alone group, and, in fact, early patients continued to receive heparin at low doses for up to 1 week after transplant. We did find a high incidence of bleeding, however, which may have also led to clots forming around the graft with subsequent infection. Since that time, in the pancreas alone group, a single loading dose of heparin is administered intraoperatively, which is not reversed, and no further heparin is given. Our series is still small; there- fore, no conclusions can be drawn yet. Some of the vascular thromboses that were believed to be technical in na- ture may have been associated with high vascular resistance in the grafts, perhaps due to edema or severe acute rejection.

We did not include a comparison with patients not requiring surgical intervention, but the populations were similar. We did not identify any other characteristics, other than a prior kidney transplant, that distinguished this group from the other patients. The duodenal segment leaks did not appear to be secondary to rejection but instead to devascularization. One patient was noted to have cytomegalovirus infection in the duodenal segment, which caused the perforation.

Most complications of diabetes have stabilized after pancreas transplantation. Improvements have been seen in neuropathy and gastroparesis after pancreas transplantation, but most complications of diabetes have only stabilized.

THE AMERICAN JOURNAL OF SURGERY VOLUME 164 NOVEMBER 1992 551

surgical complications in solitary pancreas and combined pancreas-kidney transplantations

Documents