Management of the Failing or Thrombosed Hemodialysis Access
William A. Marston
Robert Mendes
An autogenous access in the upper extremity is currently the optimal access for chronic hemodialysis. Once mature, these may function for years with a low rate of complications. Despite increases in the prevalence of autogenous accesses over the past several years, the majority of patients across the country are dialyzed through prosthetic accesses. Indeed, the Dialysis Outcomes and Practice Patterns Study reported that 58% of the patients in the United States are dialyzed through prosthetic accesses, while only 24% are dialyzed through autogenous accesses, with the balance comprised of tunneled catheters.
Although the prosthetic accesses usually function well initially, their long-term patency rates are poor, with the mechanism of failure primarily related to the development of intimal hyperplasia at the venous anastomosis. Unfortunately, these access failures represent a significant burden on the health care system. The average hemodialysis patient can expect that his or her prosthetic access will fail (thrombose) every 12 to 15 months, and access-related complications are the leading cause of admission for hemodialysis patients, accounting for more than $500 million per year in health care costs.
Management of the failing and thrombosed hemodialysis access is an integral component of the care of all hemodialysis patients, and it is imperative that each access surgeon be well versed in the various treatment algorithms. In this chapter, we will review the open, surgical, and endovascular techniques for managing the failing and thrombosed hemodialysis accesses, with emphasis on prosthetic accesses. In most practice situations, a combination of these techniques will provide optimal results, and they should be viewed as complementary.
Diagnostic Considerations
The diagnosis of a thrombosed hemodialysis access is fairly straightforward and often made by the patient or the dialysis center technologists at the time of the patient’s scheduled dialysis treatment. Occasionally, it can be difficult to determine whether a prosthetic access is thrombosed in obese patients if it is tunneled relatively deep to the skin. Duplex ultrasound may be helpful to confirm the diagnosis in this setting.
A potential failing access can be identified by any number of means, including physical examination, elevated venous pressures during dialysis, abnormal urea/recirculation measurements, unexplained decreases in measurement of the dialysis dose, or changes in access flow. The specific method of detecting the failing access is contingent upon the preference of the individual dialysis center. However, it is imperative that every access be examined by the dialysis technologist during each dialysis treatment and that each center adopts a formal surveillance protocol as recommended by National Kidney Foundation Dialysis Outcome Quality Initiative Guidelines (K/DOQI). A variety of protocols have been developed to survey prosthetic accesses, although they are not very well defined for autogenous accesses. It has been our anecdotal impression that autogenous accesses usually present as failing ones long before they thrombose and are, therefore, amenable to intervention, similar to the scenario with failing lower-extremity bypasses.
Pathogenesis
The majority of prosthetic accesses fail due to the development of intimal hyperplasia at the venous anastomosis and/or venous outflow tract. The causes of prosthetic access failure in our recent clinical experience are shown in Table 85-1. Notably, venous outflow problems accounted for 85% of the failures with 55% of the lesions limited to the venous anastomosis (Fig. 85-1) and 30% due to more extensive, long-segment stenoses or outflow occlusions (Fig. 85-2). Multiple other reports in the literature have supported our findings and have emphasized the significance of the venous anastomosis and outflow tract as the etiology of prosthetic access failure. As shown in our experience, stenoses at the arterial anastomosis and within the prosthetic graft itself can contribute but are clearly secondary.
The mechanisms responsible for the failure of autogenous access are not as well described. Hemodynamically significant lesions secondary to intimal hyperplasia can develop within the autogenous access and the ipsilateral central veins and clearly can contribute to their failure. The specific location of these offending lesions is not as consistent as associated with prosthetic accesses, but they frequently occur at the arterial anastomosis and within the proximal few centimeters of the access. The natural history of autogenous and prosthetic accesses are likely different, with the former far more resistant to thrombosis, presumably
due to the antithrombotic properties of the vessel wall.
due to the antithrombotic properties of the vessel wall.
Table 85-1 Identifiable Causes of Prosthetic Access Thrombosis (N = 115) | |||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
The prosthetic and autogenous accesses are prone to develop pseudoaneurysms (prosthetic) and true aneurysms (autogenous). These lesions may cause the access to thrombose, as with the case of an anastomotic pseudoaneurysm after an aortobifemoral bypass. However, the greater concern is that the aneurysm/pseudoaneurysm may erode through the skin and cause significant hemorrhage. Fortunately, this is a relatively rare event. Prosthetic access pseudoaneurysms result from the degeneration of the prosthetic material itself and usually result from repeated cannulation in the same segment of the graft. Aneurysms in an autogenous access result from the continued dilation of the vein that comprises the access itself. These aneurysms likely result from the same hemodynamic forces that caused the vein to dilate initially, although it has been proposed that repeated cannulations in the same segment of the autogenous access may lead to the aneurysmal degeneration similar to the situation with prosthetic pseudoaneurysms. Furthermore, it has been our anecdotal impression that autogenous accesses that have an aneurysmal segment frequently have a hemodynamically significant stenosis in the venous outflow tract.
 Figure 85-1. A fistulagram/venogram of a forearm prosthetic access is shown demonstrating a high-grade stenosis of the venous anastomosis. |
Indications and Contraindications
Given the limited number of access sites available for each patient and the increasing life expectancy of patients on hemodialysis, it is desirable to extend the lifespan of each access as long as possible. Within this context, all failing and thrombosed accesses should be corrected and/or salvaged if possible. This should be performed expeditiously in the outpatient setting using local anesthesia, and the use of temporary hemodialysis catheters should be avoided. There are both open, surgical, and endovascular treatment options available for treating the failing and thrombosed access, although no consensus exists as to the optimal approach. Notably, Green et al. recently performed a meta-analysis of the seven randomized trials comparing open, surgical, and endovascular treatment of thrombosed prosthetic accesses. They concluded that the patency rates associated with open, surgical treatment were superior for every time point analyzed. Despite these findings, the endovascular or percutaneous approach affords many advantages, including the fact that it is relatively simple, less invasive, well tolerated form a patient perspective, and can be performed in an imaging suite and, therefore, does not necessarily require the operating room. Importantly, the open, surgical, and endovascular treatments should likely be viewed as alternative or complementary approaches rather then competitive ones.
K/DOQI has carefully outlined the treatment of failing and thrombosed accesses and has defined performance standards. They recommend that all hemodynamically significant stenoses (>50%) associated with clinical or physiologic abnormalities should be corrected using either open, surgical, or endovascular techniques. Notably, Lumsden et al. reported from a randomized, controlled trial that prophylactic balloon angioplasty of venous outflow stenoses greater than 50% did not improve the patency rates of prosthetic accesses. The explanation for the differences between these Level 1 findings and the K/DOQI recommendation is not clear, but it may be due to the definition in K/DOQI of a hemodynamically significant stenosis with a “clinical or physiologic abnormality.” Furthermore, K/DOQI recommend that all thrombosed prosthetic accesses should be corrected with either open, surgical, or endovascular-based mechanical/pharmacomechanical means. They state that the success rate for clearing a thrombosed autogenous access is poor, and they defer management to the individual institution. K/DOQI state that the unassisted patency rates after open, surgical, and endovascular treatment of a failing prosthetic access should be 50% at 1 year and 50% at 6 months, respectively. The patency goal after endovascular salvage of a thrombosed prosthetic access is 40% at 3 months, while that for open, surgical salvage is 50% at 6 months. The K/DOQI justify the higher standards for the open, surgical procedures, because they are more invasive and may use the outflow veins that extend more proximally on the arm. Unfortunately, few prospective studies examining the role of endovascular or open, surgical treatment of thrombosed prosthetic accesses have matched the K/DOQI performance standards.
Regardless of the initial treatment (open surgical vs. endovascular) for the thrombosed
access, it is imperative to identify and correct the underlying cause of the failure, if at all possible. This translates into correcting the venous outflow stenosis in the majority of cases. Multiple previous studies have shown that simply removing the thrombus within the access alone is insufficient and does not result in long-term patency. Indeed, K/DOQI recommend that a completion fistulagram should be performed after lysis of the clot and the residual stenoses corrected. Selecting the appropriate remedial therapy requires an understanding of the long-term success rates of the procedures for the specific lesion. In a study of 59 prosthetic grafts surgically revised, we reported a significant variability in the long-term access success based upon the offending cause. The functional patency rate was 44% at 6 months for patients with venous anastomotic stenoses but only 18% for those patients with diffuse venous outflow stenoses. Occasionally, the completion fistulagram may indicate that attempts at salvage are unlikely to be successful. In this setting, further salvage efforts should be abandoned and a completely new access constructed.
access, it is imperative to identify and correct the underlying cause of the failure, if at all possible. This translates into correcting the venous outflow stenosis in the majority of cases. Multiple previous studies have shown that simply removing the thrombus within the access alone is insufficient and does not result in long-term patency. Indeed, K/DOQI recommend that a completion fistulagram should be performed after lysis of the clot and the residual stenoses corrected. Selecting the appropriate remedial therapy requires an understanding of the long-term success rates of the procedures for the specific lesion. In a study of 59 prosthetic grafts surgically revised, we reported a significant variability in the long-term access success based upon the offending cause. The functional patency rate was 44% at 6 months for patients with venous anastomotic stenoses but only 18% for those patients with diffuse venous outflow stenoses. Occasionally, the completion fistulagram may indicate that attempts at salvage are unlikely to be successful. In this setting, further salvage efforts should be abandoned and a completely new access constructed.
 Figure 85-2. A venogram of the outflow tract of a prosthetic access is shown. The angiographic catheter is positioned in the axillary vein. Note the diffuse stenosis within the axillary and subclavian veins that comprise the outflow tract, in addition to the large, proximal collateral vein. |
Both pseudoaneurysms and true aneurysms associated with prosthetic and autogenous accesses should be corrected if the overlying skin is threatened due to their potential to rupture. Furthermore, all infected prosthetic pseudoaneurysms should be corrected. Notably, the K/DOQI recommend that prosthetic pseudoaneurysms that expand rapidly and those that measure twice the size of the graft should be repaired, although they state that only the autogenous aneurysms that involve the arterial anastomosis merit repair.
Pre-operative Assessment
The extent of the pre-operative assessment is dictated by the planned procedure. The endovascular procedures require only a minimal pre-operative evaluation, while that for the open, surgical procedures is identical to that for the initial access procedure and comparable to that for most vascular surgical procedures. It is imperative to determine whether the patients need to dialyze prior to any planned procedure, because they frequently present to the dialysis unit with a thrombosed graft and thereby miss their planned session. This can usually be determined by an assessment of their volume status and serum electrolytes. Consultation with the patient’s attending nephrologist may further assist with this determination. It is usually possible to cannulate the femoral vein with a temporary catheter for patients that need to be dialyzed emergently prior to any intervention to salvage their permanent access.
Operative Technique
Endovascular Treatment for Failing and Thrombosed Prosthetic Accesses
The endovascular treatment of failing or thrombosed prosthetic accesses is commonplace given their overall prevalence and poor long-term patency rates. The relative breakdown of interventions for failing or thrombosed accesses is contingent upon the presence of a routine screening program at the respective dialysis centers. The management of the failing and thrombosed prosthetic access is similar in terms of the diagnostic and therapeutic interventions, but it will be discussed separately.
Failing Prosthetic Accesses
Patients with a failing prosthetic access are typically referred to the access surgeon with a history of poor access flow, increased outflow pressures, or poor clearance rates, although the specific functional abnormality is contingent upon the surveillance method used in their center. The objective of the initial diagnostic procedure is to identify the underlying problem. As noted above, the majority of the problems are localized to the venous outflow site, but other lesions contribute in approximately 20% of the cases. The extent of the initial diagnostic procedure is dictated by the clinical suspicion of the underlying problem (i.e., arterial inflow, venous outflow); diagnosis with a fistulagram that visualizes the arterial anastomosis, prosthetic access, and complete venous outflow is usually sufficient. In the rare instances in which an arterial inflow problem is suspected, a formal upper-extremity arteriogram from the aortic arch to the access arterial anastomosis may be required in addition to the fistulagram/venogram.
The prosthetic access is punctured approximately 5 cm from the arterial anastomosis with the needle directed toward the venous anastomosis. We prefer to use either a straight angiographic needle and 0.035 in. wire or a micropuncture system with a 0.018 in wire. Local anesthesia is used for the puncture site, and intravenous sedation is administered as necessary to assure patient comfort. It is important to select the proper puncture site to allow sufficient working room for any interventional procedure. Because stenoses at the venous anastomosis is the most likely cause of access failure, the puncture site must not be too close to this area. A sheath is then placed over the initial access wire with the size contingent upon the puncture system used. We routinely use a 4 French sheath with the angiographic needle and a 3 French sheath for the micropuncture system. Regardless of its size, the sheath tip should be seated proximal to the venous outflow tract of the access to assure adequate visualization of the venous anastomosis. A guidewire is then carefully advanced across the venous anastomosis and extended into the venous outflow tract (Fig. 85-3). The specific wire is somewhat operator dependent with either a starter (e.g., Bentson) or selective (e.g., Glidewire) wire sufficient. Serial digital subtraction images of the access, including the arterial anastomosis and complete venous outflow, are then obtained using manual injection of the contrast. Importantly, the venous outflow tract should include the axillary, subclavian, and brachiocephalic veins in addition to the superior vena cava (Fig. 85-4A and 85-B). A catheter may be advanced into the central outflow veins over the guidewire to permit localized injection of the contrast. Occasionally, visualization of the venous outflow tract is poor due to vasospasm. In these cases, vasodilators such as nitroglycerin or papaverine can be used. The visualization of the arterial inflow and anastomosis can be facilitated by manually compressing the venous anastomosis and/or venous outflow in an attempt to reflux the contrast retrograde.
 Get Clinical Tree app for offline access 
|
