Abstract
Vascular complications after pancreatic transplantation carry a high rate of graft loss. Endovascular management of these complications is confined to stent placement for iliac artery inflow disease and embolization for arteriovenous fistulae (AVFs), pseudoaneurysms, or active bleeding. The current study describes the endovascular management of pancreatic transplant venous thrombosis (N = 1), arterial stenosis (N = 5), thrombosis (N = 3), pseudoaneurysms (N = 1), and AVF (N = 2). In addition, embolization of nonfunctioning grafts is described as an endovascular alternative to pancreatectomy.
Keywords
Introduction
Pancreatic transplantation is a definitive treatment for diabetes and is increasingly being utilized to treat type-1 diabetes mellitus. 1–2 Pancreatic transplantation is a technically challenging surgery and vascular complications are the most common cause of early graft failure. 2–3 Vascular complications represent more than 50% of early (within 6 months from transplant) and less than 10% of late (after 6 months from transplant) pancreatic graft failure. 4
Reports of endovascular management of pancreatic transplant vascular complications are scant and in the form of case reports. 5 –13 These reports are confined to stent placement for iliac artery inflow disease 5–6 and embolization for arteriovenous fistulae (AVF), 7–8 pseudoaneurysms, or active bleeding. 10 –13
The current study describes and reports the technical and clinical outcomes of the endovascular management of a broader set of vascular complications following pancreatic transplantation. In addition, endoluminal embolization of nonfunctioning grafts is also described as an endovascular alternative to surgical pancreatectomy.
Materials and Methods
Study Design
This is a retrospective study auditing radiological examinations, charts, and records of adult pancreatic transplant recipients from January 1996 to December 2010 (15 years). Institutional internal review approval was obtained categorizing the study as secondary use of preexisting anonymous data.
Demographic information collected on pancreatic transplant recipients included age and gender. In addition, the time lapse between the pancreatic transplant and the diagnosis of the vascular complication was noted along with the diagnostic modality. The arterial vascular surgical anatomy of the current study is shown in Figure 1. Vascular complications were mostly diagnosed with contrast-enhanced magnetic resonance (MR) angiography (Figure 2A) with or without Doppler ultrasound and/or contrast-enhanced computed tomography. Vascular complications intended for endovascular interventions underwent a diagnostic angiogram prior to the intervention (Figure 2B). Imaging follow-up, if any, was done by contrast-enhanced 3-dimensional MR angiography. 2
Postpancreatic transplant vascular surgical anatomy. Pancreatic transplant arterial supply is created using a donor iliac Y-graft (Y-G). The singular limb (base) of the Y-graft is anastomosed (Prox Anst: Proximal anastomosis) end-to-side to the recipient aorta, common (CIA) or external (EIA) iliac artery. One limb of the Y-graft is anastomosed (SMA-A: SMA anastomosis) to the graft SMA. The other limb is anastomosed (SpA-A: SpA anastomsis) to the splenic artery (SpA). The gastroduodenal artery (GDA), off the common hepatic artery, communicates with the inferior pancreaticoduodenal artery (IPDA) which comes off the SMA. The IPDA and GDA form the pancreaticoduodenal arcade (PDA) supplying the pancreatic head (H). The splenic artery supplies the majority of the pancreatic body (B) and tail (T). Venous allograft drainage (not shown) is into the splenic and mesenteric vein(s) which drain anatomically into the graft portal vein. The allograft portal vein (at the authors institution) is anastomosed systemically to the recipient inferior vena cava or iliac vein. Ao indicates aorta; CIA, common iliac artery; EIA, external iliac artery; SmB, small bowel; SMA, superior mesenteric artery (pancreatic duct communication/anastomosis with the small bowel is not shown).
Pancreatic graft artery stenosis at the distal SMA anastomis with subsequent angioplasty. A, Magnetic resonance angiogram (MRA) demonstrating a stenosis (solid white arrow) of one of the distal anastomoses between the Y-graft and the superior mesenteric artery (SMA). The proximal Y-graft anastomosis (between dashed arrows) with the common iliac artery is patent. The SMA defines the border between the pancreatic head (H) and body (B). B, Selective pancreatic graft angiogram demonstrating a stenosis (arrow) of one of the distal anastomoses between the Y-graft and the SMA. The other distal anastomosis is between the Y-graft and the splenic artery (SpA). The proximal anastomosis is between the base of the Y-graft and the external iliac artery (EIA). The pancreaticoduodenal artery (PDA) is seen coming off the SMA. C and D, Two images of a digitally subtracted angiogram in sequence demonstrating improvement of the distal SMA anastomotic stenosis with residual stenosis less than 50% relative to the SMA and less than 30% relative to the Y-graft (Y). There is improved flow into the pancreaticoduodenal artery (white arrows). The 0.014-inch wire used to angioplasty is looped (dashed wire) in the SMA blind loop.
Graft function was determined by insulin independence and normal serum fasting blood sugar. 4 Graft loss was considered when at least one of the following occurred: loss of insulin independence, 4 surgical graft pancreatectomy, intentional endovascular pancreatic artery embolization/obliteration, or patient death.
Endovascular Management of Vascular Complications
Management of occlusive complications (stenoses and thromboses) was by anticoagulation for all patients and endovascular procedures in selective cases. Endovascular procedures were categorized as (1) embolization procedures, (2) angioplasty with or without stent placement procedures, and (3) transcatheter thrombolysis procedures. Embolization procedures were performed with standard coaxial techniques utilizing coils, detachable balloons, and/or Amplatzer vascular plugs (AGA Medical Corp, Golden Valley, Minnesota). Angioplasty with or without stent placement procedures was performed with standard coaxial techniques. Intravenous heparin was used intraprocedurally (3000-6000 units) and oral anticoagulants were administered postprocedurally (coumadine to therapeutic levels). Thrombolysis procedures for pancreatic artery thrombosis (PAT) involved clot maceration utilizing a microwire and microcatheter. In addition, manual pulse spray pharmacolysis with 2 mg of tissue Plasminogen Activator (tPA) in 8 mL of normal saline (0.9%) was administered within 4 to 5 minutes via a microcatheter directly into the thrombosed graft artery. No indwelling (overnight) catheter directed pharmacolysis was utilized.
Results
Over a 15-year period, 160 pancreatic transplants were performed on 153 adult patients (7 patients underwent 2 pancreatic transplant procedures), comprising a total of 99 (65%) male patients and 54 (35%) female patients. Of the retransplants, 4 were male and 3 were female. Mean recipient age was 40 years (range: 18-62 years). Overall, 107 (67%) transplants had no vascular complications and 53 (33%) had vascular complications. The vascular complications were based on imaging and not all were clinically evident. Of the 53 vascular complications, 48 (90.5%) were diagnosed with magnetic resonance angiogram, 3 (5.7%) diagnosed with Doppler ultrasound initially, 2 (3.8%) diagnosed with CTA, and 1 (1.9%) originally diagnosed by angiography. Twenty (37.7%) of the 53 vascular complications had confirmatory angiography.
Endovascular Interventions
Endovascular procedures were performed for 12 independent vascular complications on 11 pancreatic transplants (11 patients). The 12 vascular complications treated represent 22.6% of vascular complications (n = 12/53) in 11 transplants (N = 11/160; 6.9% transplants). Table 1 demonstrates the demographics of the 12 endovascularly treated vascular complications. Table 2 (Figures 2 –4) demonstrates the vascular complications, endovascular management, and outcome of the 12 vascular complications in the 11 pancreatic transplants.
Demographics of Patients with Post-pancreatic Transplant Vascular Complications who Underwent Endovascular Therapy.
Abbreviations: PTx, pancreatic transplant; M/F, male/female; PAT, pancreatic artery thrombosis; PAS, pancreatic artery stenosis; PsA, pseudoaneurysm; AVF, arteriovenous fistula; SMA, superior mesenteric artery of graft; SPA, splenic artery of graft; IPDA, inferior pancreaticoduodenal artery; SMV, superior mesenteric vein of graft; SVT, splenic vein thrombosis; PV, portal vein of graft.
a Suspected initially by CTA and then confirmed by the second modality (MRA).
Endovascular Management and Outcome of Vascular Complications
Abbreviations: NS, normal (0.9%) saline; Mo, month/months.
a Last follow-up is the date of graft failure. Entry without superscript alphabets is the date of last follow-up with a functioning pancreatic graft.
b Graft already lost or dysfunctional. Intent was to regain function, however, that was never achieved.
Superior mesenteric to portal venous fistula embolization. A, Graft splenic artery (SpA) angiogram from a 5F vertebral catheter (arrow) with little reflux of contrast in the adjacent superior mesenteric artery (SMA). The reflux demonstrates the arteriovenous fistula (between dashed arrows) between the SMA stump and the adjacent graft portal vein. B, Selective graft SMA angiogram from the 5F Vertebral catheter (arrow) again demonstrating the arteriovenous fistula (between dashed arrows) with an early draining portal vein (PV) that has a varix (asterisk). A hint of the external iliac vein is seen. The pancreaticodoudenal artery and its branches (arrow heads) are seen. C, Selective subtracted pancreatic graft angiogram after balloon test occlusion of the arteriovenous fistula (arrow at occlusion balloon). The test occlusion helps plan the precise site of the embolization to test for hemodynamics and pancreaticoduodenal artery sparing (dashed arrows). The bracket points to the splenic artery branches supplying the pancreatic graft body and tail. D, Fluoroscopic image after microcoils has been deployed as a nest entrapped in the proximal Amplatzer plug (between hollow arrows). This proceeded impaling a microcatheter into the proximal Amplatzer plug with a 5F catheter tip abutting the proximal Amplatzer for support. E, Completion subtracted pancreatic graft angiogram after the second session of the arteriovenous embolization demonstrating complete obliteration of the fistula at the level of the previously placed Amplatzers (asterisks). The pancreaticodeodenal artery (arrows) off the superior mesenteric artery is spared.
Splenic vein transcatheter mechanical thrombectomy. A, Digital subtraction venogram through a 5F catheter in the same patient as Figure 3 (Amplatzers between dashed arrows) demonstrating partial thrombus (filling defect between solid arrows) in the splenic vein (SpV). The draining portal vein (PV) and external iliac vein (EIV) are free of clot. B, Digital subtraction venogram after mechanical thrombectomy with the Arrow-Trerotola device demonstrating residual thrombus (solid arrows) in the SpV. Again seen is the draining PV and EIV are free of clot.
In addition to the endoluminal procedures for the 12 vascular complications (Table 2), a pancreatic artery embolization was performed for complete pancreatic graft death (“endovascular pancreatectomy” procedure, authors' term). The graft suffered from fungal pancreatitis 56.9 months after transplantation. The condition responded to medical treatment, however, graft loss was diagnosed 2.7 months after the diagnosis of fungal pancreatitis had resolved. The patient underwent a Y-graft coil embolization and a balloon expandable stent-graft placement (iCast; Atrium, Hudson, New Hampshire) in the iliac artery across the transplant anastomosis. The stent-graft (iCast, Atrium) component of the procedure was performed in order to occlude the Y-graft stump and protect both stump and adjacent iliac artery from pancreatitis lead pseudoaneurysm formation. The endovascular pancreatectomy procedure was performed in an attempt to spare surgical pancreatectomy and to avoid “run-away” pancreatitis (intent of the procedure was to cause excretory graft death in addition to the already established hormonal graft loss) in a failed graft which may lead to pelvic pain, pseudocysts, and/or pseudoaneurysms (as was the case in the pseudoaneurysm embolization attempt which is outlined in Tables 1 and 2).
Discussion
The role of endovascular interventions for the management of vascular complications of pancreatic transplants is not well established due to the rarity of the procedures and the limited literature on the subject mainly consisting of anecdotal case reports describing technically successful procedures. 5 –13 Despite the fact that this study describes the broader collection of endovascular procedures in this setting, our experience is still anecdotal. The sample size is small and not all complications were treated by endovascular means. For example, over 15 years of this retrospective audit, some Pancreatic artery stenosis (PAS) were treated by endovascular methods in an ad hoc manner and some significant PAS were not. This was at the sole discretion of the attending transplant surgeons. Thus, due to the retrospective nature of the study, the small sample size, and the nonuniform management approaches; it is difficult to determine with certainty that endovascular interventions for PAS adds to the management of PAS compared with anticoagulation alone.
However, despite our limited experience, endovascular management of PAT appears to be generally unable to restore graft function. As a result, transcatheter thrombolysis should be utilized judiciously (utilized on case-by-case basis). This is similar to the experience for the pharmacolysis and mechanical lysis of thrombosed hepatic arteries in liver transplant recipients. 14 –16 Pancreatic graft arteries are generally too small and too tortuous to allow use a transcatheter mechanical thrombectomy device to be used or for a true (purposely designed) indwelling infusion catheter to be used for prolonged pharmacolysis. Furthermore, the age of the arterial thrombus is not accurately known. There may be a clinically occult period (of unknown length) prior to symptoms that would alert physicians to evaluate and potentially treat the arterial thrombus (the thrombus may be a lot older than the symptoms).
Embolization, on the other hand, if technically possible appears to be a reasonable option for the management of AVF and pseudoaneurysms presenting with graft dysfunction and bleeding or impending bleeding, respectively. This is said despite our technical failure to exclude a pseudoaneurysm. The finding of a fistulous (AVF) component with the psuedoaneurysm has been noted by prior authors and is not a surprise in the setting of pancreatitis and/or iatrogenic trauma. 5
From a technical standpoint, this study also describes an endovascular bailout technique helpful in cases of Amplatzer vascular plug failure (AGA Medical Corp, Golden Valley, Minnesota) regardless of the target vessel. If there is no room to place additional Amplatzer devices or coils, as in the case in this study (Figure 3D and E), impaling the Amplatzer plug with a microcatheter system which is coaxially placed and supported by a 4- or 5-F catheter abutting the Amplatzer subsequently followed by coil embolization within the Amplatzer seems to be effective. The Amplatzer (AGA Medical Corp) acts as nester or cage that prevents coil maldeployment.
The current study describes a new procedure which may be the endovascular equivalent of surgical graft pancreatectomy. In an attempt to spare surgical pancreatectomy and to avoid "run-away" pancreatitis in hormonally failed grafts (which may lead to pseudocysts and/or pseudoaneurysms), transcatheter arterial embolization with/or without iliac stent graft placement can be performed in an attempt to establish excretory graft death in addition to the already established hormonal graft loss. However, endovascular graft embolization may only be an alternative to graft pancreatectomy in grafts with patent arteries (no reason and probably not technically feasible to embolize in the setting of complete PAT) and in the absence of clinical signs of graft infection.
In conclusion, endovascular management of AVF and possibly PAS are technically feasible and appear to be clinically effective, however, they are not effective for PAT in our small series. Endovascular embolization for induction of excretory pancreatic death after hormonal graft loss is a potential alternative for graft explantation and needs to be studied further regarding feasibility, clinical applications, and outcome.
Footnotes
The author(s) declared no conflicts of interest with respect to the research, authorship, and/or publication of this article.
The author(s) received no financial support for the research, authorship, and/or publication of this article.