Pancreas transplantation can restore normal glycemic control and thus has the potential to cure type 1 diabetes.1 Because of the need for lifelong immunosuppression and the morbidity and mortality associated with major surgery, pancreas transplantation is only performed in a selection of patients with type 1 diabetes: in combination with a kidney transplant in patients with end-stage renal disease and as a solitary transplant in patients with untreatable acute diabetes complications such as frequent severe hypoglycemia.1
The first simultaneous pancreas-kidney (SPK) transplant was done by Kelly et al in 1967.2 As surgical techniques and immunosuppressive regimens have evolved, patient and graft survival have improved.3 In the early days, most SPK transplantations were performed using bladder drainage.4 Because of the high rate of urological complications associated with bladder drainage, enteric derivation has been increasingly performed as preferential exocrine derivation.5-7
In SPK transplantation, complications can be related to the renal graft, the pancreatic graft, or can be general surgical complications.8 Most complications, especially the serious ones, are related to the pancreatic graft, such as graft pancreatitis, intra-abdominal infection, graft thrombosis, and bleeding.8 Pancreatic graft-related complications can also arise from the simultaneously transplanted duodenum.5,9 The majority of all complications described in the literature are early complications, usually in the immediate postoperative period.5,8,10-12 In this report, we present 3 cases of bleeding from the transplant duodenum occurring several years after transplantation.
A 58-year-old female with type 1 diabetes underwent SPK transplantation in 2006 because of end-stage diabetic nephropathy (Table 1, patient 1). In March 2017, she was diagnosed with low-grade serous ovarium carcinoma (FIGO stage IIb), and treatment consisted of hysterectomy, bilateral salpingo-oophorectomy, and omentectomy 3 months later. Subsequently, adjuvant treatment with letrozole was administered.
In August 2017, 10 years after transplantation, the patient presented with fatigue and atypical chest pain. She was known to have a decreased left ventricular systolic function after a postoperative anterolateral myocardial infarction in 2006, but troponins on admission were normal, electrocardiography was unchanged, and myocardial perfusion scintigraphy was normal. Lab results showed anemia with an important drop in hemoglobin level (hemoglobin 6.0 g/dL, as compared with 8.2 g/dL 12 days before). Baseline hemoglobin was within the normal range until 1 year before presentation. Since August 2016, a slow progressive decline in hemoglobin with iron deficiency was noted. Vitamin B12 and folic acid concentrations were normal. Prothrombin time was normal, and a mild leukopenia and discretely diminished platelet count were present, additionally. The patient was taking low-dose aspirin since 2006 as secondary prevention after myocardial infarction. Gastroscopy showed ulcers in the distal part of the bulbus of the duodenum without signs of bleeding. A proton-pump inhibitor was started. Mid and end of September 2017, she was readmitted twice because of melena with a concomitant drop in hemoglobin. A repeat gastroscopy showed healed ulcers, and colonoscopy could not reveal a bleeding focus. Computed tomography (CT) angiography demonstrated enlarged and tortuous veins of the transplant duodenum at the site of anastomosis, but without signs of active bleeding and without signs of venous thrombosis. Because of persistent discrete pancytopenia with recurrent melena, immunosuppression with azathioprine and tacrolimus was switched to methylprednisolone and tacrolimus. Capsule endoscopy showed hyperemic and congestive mucosal folds with a small amount of blood in the entero-enteric anastomosis (Figure 1). The initial anastomosis in 2006 was a side-to-side duodenojejunostomy (with systemic drainage). Reintervention with Roux-en-Y reconstruction of the anastomosis was performed, and a congestive and very friable mucosa was noted during surgery. Anatomopathological examination of the resection specimen showed discrete ischemic damage of the mucosa, and submucosal veins were enlarged and congestive (Figure 2). Postoperatively, the patient developed pancreatitis with a temporary need for insulin therapy. On day 2 postoperative, low-dose aspirin was restarted. During a follow-up period of 6 months, there was no recurrence of melena, hemoglobin increased spontaneously, and there was no need for transfusion anymore.
A 57-year-old female with type 1 diabetes underwent SPK transplantation in 1993, because of end-stage diabetic nephropathy (Table 1, patient 2). Initial transplantation was performed with bladder drainage. In 2003, during explorative laparotomy for a suspicious adnexal mass, a small opening in the donor duodenum was noted, and conversion to enteric drainage with side-to-side duodenojejunostomy with systemic drainage was made during the same procedure. The patient presented in April 2015, 22 years after transplantation, with melena with a significant drop in hemoglobin (Hb 8.3 g/dL, hemoglobin 1 mo earlier was 12.7 g/dL). There was no thrombocytopenia, and prothrombin time was normal. The patient was on dual antiplatelet therapy (aspirin and clopidogrel) for several years because of a past medical history of acute myocardial infarction, coronary artery bypass grafting, and percutaneous transluminal angioplasty of the femoral artery. Gastroscopy showed a hiatal hernia with a Cameron lesion without signs of active bleeding (a proton-pump inhibitor was started), and colonoscopy was completely normal. Dual platelet therapy was temporarily interrupted, and aspirin monotherapy was restarted a few days later without signs of recurrent bleeding. One month later, in May 2015, the patient was readmitted because of recurrent melena with a concomitant drop in hemoglobin (Hb 6.5 g/dL, hemoglobin 5 days earlier was 9.3 g/dL). CT angiography did not show any extravasation of contrast. However, capsule endoscopy showed a small active bleeding in the duodenum of the donor (Figure 3). The lesion was not accessible endoscopically, and the melena stopped spontaneously, so the patient was discharged again. One month later, in June 2015, she represented with the same symptoms (Hb 6.9 g/dL, hemoglobin 6 days earlier was 11 g/dL), but this time an active bleeding of the jejunal branch of the superior mesenteric artery was seen on CT angiography. The lesion was coiled by the interventional radiologist. Aspirin was temporarily interrupted, and the initial evolution was satisfactory; however, in September 2015, she again presented with melena, although hemoglobin was stable (Hb 9.2 g/dL, as compared with 9.2 g/dL 1 wk before). At this time, it was decided to perform a reintervention with a Roux-en-Y reconstruction of the anastomosis. Perioperatively, a congestive and very friable mucosa was noted but without signs of active bleeding. Histologic examination of the resection specimen showed slight congestion of the submucosal veins. As in our first patient, there was no recurrence of melena or hemoglobin level drop during a follow-up period of 2 years and 7 months, despite restarting aspirin in a low dose.
A 65-year-old male with type 1 diabetes received an SPK transplantation with bladder derivation in 1999 because of end-stage diabetic nephropathy (Table 1, patient 3). Since 2014, he restarted renal replacement therapy because of transplant kidney failure. In December 2017, a coronarography was done as part of a cardiovascular work-up in preparation for second kidney transplantation. Coronarography revealed significant stenoses of the right coronary artery and the mid left anterior descending and the distal left anterior descending coronary artery. Percutaneous coronary intervention with placement of 5 drug-eluting stents was necessary, and dual antiplatelet therapy (aspirin and clopidogrel) was started.
The patient presented 1 month later, in January 2018, 18 years after transplantation, with severe hematuria with hemodynamic instability and needed multiple transfusions (Hb 8.8 g/dL, baseline hemoglobin 12.0 g/dL 1 mo before). Platelets and prothrombin time were normal. CT angiography did not reveal any active bleeding, and cystoscopy did not reveal hemorrhagic cystitis. Renal MR was able to exclude underlying malignancy or other bleeding focus in the kidneys. By exclusion, we assumed that the bleeding originated from the transplant duodenum. Aspirin was stopped, and clopidogrel was continued in monotherapy. The first episode of hematuria ended spontaneously, but there was a quick recurrence of hematuria. Surgical revision with enteric derivation of the pancreas was planned, but perioperative findings showed an extremely fragile intestinal tissue with a high risk of leakage, bleeding, or loss of transplant pancreas. For this reason, enteric derivation was not possible. Until today, after 2 months of follow-up, there was no recurrence of hematuria.
In this case series, we present 3 cases, 2 enterically and 1 bladder-derived kidney-pancreas transplant patients, with anastomotic hemorrhage as a very late complication of SPK (11, 22, and 18 y later, respectively). Anastomotic hemorrhage between the jejunum and duodenal stump is a known early complication after pancreas transplantation with enteric derivation. In the study of Orsenigo et al, with 61 enteric-drained kidney-pancreas transplant recipients, 11% of patients experienced bleeding complications in the first week after transplantation.5 Anastomotic ulcers between the duodenal stump and the recipient duodenum were the most common site of hemorrhage in this study.5 In bladder-drained pancreas transplants, hematuria can occur as a result of anastomotic hemorrhage with or without ulcerative lesions of the anastomosis.4
Delayed bleeding of the duodenal anastomosis has only rarely been reported in literature. Pieroni et al9 recently reported in their series of 336 pancreas transplant patients that 4.5% of patients developed delayed duodenal graft complications requiring duodenectomy, of which 28% were duodenal graft bleedings. Bleeding occurred between 2 and 50 months, with a median of 44 months9 (Table 2). Not a single pancreas graft loss occurred as a consequence of delayed bleeding.9 No clear etiology could be identified. In the series of Messner et al, containing 379 pancreas transplant patients, 1.3% of patients developed late bleeding of the duodenal graft anastomosis, with a mean delay of 6.4 years after transplantation13 (Table 2). In 2 of these patients, histopathology of the donor duodenum and anastomosis showed enlarged submucosal veins; in 1 of the patients, there were signs of thrombosis with ischemic damage, and in 1 patient the resection specimen was normal (Table 2).13 As in the case series of Pieroni et al, there were no pancreatic graft losses as a consequence of bleeding, but 1 patient died 2 years after the last bleeding due to myocardial infarction.13
The 3 patients included in this case series are the only patients in our cohort presenting with delayed bleeding of the transplant duodenum. In our center, between January 1992 and June 2018, 122 SPK transplantations have been performed. This means that 2.5% of our patients developed delayed bleeding of the transplant duodenum, with a median delay of 18 years after transplantation. In our case series, like in the series of Pieroni et al and Messner et al, there were no pancreatic graft losses due to this bleeding.
In the first 2 patients, a congestive and friable mucosa was seen perioperatively, and examination of the resection specimen showed enlarged and congestive submucosal veins in both patients, comparable to the findings in 2 of the patients in the series of Messner et al. In the third patient, with a bladder-derived kidney-pancreas transplantation, because of the extremely fragile intestinal tissue perioperatively, enteric derivation was not possible. This fragile intestinal tissue is the likely explanation for the episodes of bleeding but also precluded further surgical intervention because of the high risk of perioperative bleeding and consequent risk of losing the transplant pancreas. Urinary reflux was never shown in this patient, but mucosal irritation by urinary reflux could have contributed to the bleeding in this patient.
Until present, the mechanism of development of this vulnerable mucosa with enlarged, congestive submucosal veins remains unclear. A possible mechanism could be an increase in vascular resistance or partial obstruction on the venous side, creating a decreased blood flow and enabling the development of duodenal varices. If we presume this concerns a slow process, it could explain the late nature of these bleedings, several years after transplantation. In our patients, we did not find any radiological nor histological evidence of thrombosis of the venous drainage system. During the revision of the anastomosis in the first 2 patients, perioperatively, there were no signs of vascular thrombosis seen.
Other possible mechanisms of late bleeding in SPK transplant patients could be a chronic low-grade nonocclusive ischemia because of a low flow phenomenon in a population with a known high cardiovascular risk, comparable to the mechanism of development of ischemic colitis. If this low flow phenomenon develops slowly, it could again be a possible explanation for the late nature of bleeding in SPK transplant patients.
In both enterically and bladder-derived pancreas transplantation, cytomegalovirus can be a precipitating factor causing duodenal ulcers that can cause bleeding.4,5 In our first 2 patients, cytomegalovirus PCR in serum was negative; in the third patient, it was not tested, but there were no clinical or biochemical signs of CMV infection. There was no histological evidence of CMV in the resection specimens. Chronic rejection could also be a precipitating factor, but resection specimens of the first 2 patients did not show any sign of rejection, and there was no evidence of concomitant renal allograft rejection.
All of our patients were on low-dose aspirin and/or clopidogrel for secondary cardiovascular prevention. Antiplatelet therapy is frequently used in the population of patients who undergo kidney and pancreas transplantation because of their increased cardiovascular risk. Possibly it could be a precipitating factor, so we suggest discontinuation of antiplatelet therapy if possible. In a study by Benko et al, 21.6% of patients listed for kidney transplantation received oral antiplatelet medication, and among them, 10% received dual antiplatelet therapy.14 In patients receiving SPK transplantation, the numbers of patients taking antiplatelet therapy are poorly described in literature but are probably even higher because of concomitant diabetes. Dual antiplatelet medication (but not monotherapy) was associated with a higher risk for early postoperative bleeding.14 Further studies are needed to investigate if there is a link between antiplatelet therapy and late bleeding in kidney-pancreas transplant recipients.
Bleeding of the transplanted donor duodenum can present as a late complication, several years after both enterically derived and bladder-derived kidney-pancreas transplantation. The development of enlarged, congestive submucosal veins could play a role in these late bleedings, and antiplatelet therapy could be a possible precipitating factor of bleeding. Further research is necessary to investigate the pathophysiology, the prevalence, optimal treatment, and the consequent influence on mortality, morbidity, and graft loss after SPK transplantation.
1. American Diabetes Association. Pancreas and islet transplantation in type 1 diabetes. Diabetes Care. 2006; 29:935
2. Kelly WD, Lillehei RC, Merkel FK, et al. Allotransplantation of the pancreas and duodenum along with the kidney in diabetic nephropathy. Surgery. 1967; 61:827–837
3. Sutherland DE, Gruessner RW, Dunn DL, et al. Lessons learned from more than 1,000 pancreas transplants at a single institution. Ann Surg. 2001; 233:463–501
4. Hakim NS, Gruessner AC, Papalois BE, et al. Duodenal complications in bladder-drained pancreas transplantation. Surgery. 1997; 121:618–624
5. Orsenigo E, Fiorina P, Dell’Antonio G, et al. Gastrointestinal bleeding from enterically drained transplanted pancreas. Transpl Int. 2005; 18:296–302
6. Gruessner AC, Sutherland DE. Pancreas transplant outcomes for United States (US) cases reported to the United Network for Organ Sharing (UNOS) and non-US cases reported to the International Pancreas Transplant Registry (IPTR) as of October, 2000. Clin Transpl. 200045–72
7. West M, Gruessner AC, Metrakos P, et al. Conversion from bladder to enteric drainage after pancreaticoduodenal transplantations. Surgery. 1998; 124:883–893
8. Grochowiecki T, Gałązka Z, Madej K, et al. Multivariate analysis of complications after simultaneous pancreas and kidney transplantation. Transplant Proc. 2014; 46:2806–2809
9. Pieroni E, Napoli N, Lombardo C, et al. Duodenal graft complications requiring duodenectomy after pancreas and pancreas-kidney transplantation. Am J Transplant. 2018; 18:1388–1396
10. Khubutia MS, Pinchuk AV, Dmitriev IV, et al. Surgical complications after simultaneous pancreas-kidney transplantation: a single-center experience. Asian J Surg. 2016; 39:232–237
11. Chan CM, Chim TM, Leung KC, et al. Simultaneous pancreas and kidney transplantation as the standard surgical treatment for diabetes mellitus patients with end-stage renal disease. Hong Kong Med J. 2016; 22:62–69
12. Redfield RR, Rickels MR, Naji A, et al. Pancreas transplantation in the modern era. Gastroenterol Clin North Am. 2016; 45:145–166
13. Messner F, Bösmüller C, Oberhuber R, et al. Late recurrent bleeding episodes from duodenojejunostomy after pancreas transplantation. Clin Transplant. 2018; 32:e13350
14. Benkö T, Gottmann M, Radunz S, et al. One-year allograft and patient survival in renal transplant recipients receiving antiplatelet therapy at the time of transplantation. Int J Organ Transplant Med. 2018; 9:10–19