Thrombosis of arteriovenous access is related to the same issues that cause thrombosis and any other blood vessel: Virchow’s triad. Virchow outlined three causes of thrombosis in any blood vessel: an abnormality of the blood vessel, an abnormality of flow, and an abnormality in coagulation. Access thrombosis is no different. Typically, as the access is used for dialysis, cannulation of the conduit generates trauma to the vessel wall. This cannulation trauma can incite thrombosis at the site of cannulation or, more commonly, lead to false aneurysm formation.

The greatest risk to an arteriovenous access is venous outflow stenosis associated with intimal hyperplasia [5]. Often this has no relationship to cannulation whatsoever. Nonetheless, the development of venous stenosis is often a precursor before access thrombosis. This is particularly true in prosthetic access since intimal hyperplasia tends to develop at the venous end of the artificial conduit. This slows flow in the prosthetic graft leading to thrombosis of the prosthetic graft from its beginning all the way to the venous end stenosis. Management is therefore focused on relieving the venous and stenosis to normalize graft blood flow.

Rarely hypercoagulability can become an issue in a mature arteriovenous access. Generally these acquired forms of hypercoagulability are not common, but when they occur, long-term arteriovenous access can be challenging [6–9]. At this point, routine screening for hypercoagulable states is not recommended [6, 10].

Concerns relative to arteriovenous access thrombosis are raised by any evidence slowing the flow in the access or by venous hypertension in the access [11–13]. Typically, as venous outflow stenosis develops in the access, prolonged bleeding occurs at the time of the cannulation after dialysis, often associated with an increased pulsatility in the access. Prolonged bleeding is often the first sign of impending access failure but may be so subtle as to initially be missed. With time, the bleeding may persist for a longer period of time and may be more voluminous. Prolonged bleeding should lead to referral for access evaluation and management.

Sometimes, external bleeding is not the issue. In many cases hematoma formation may develop after decannulation, potentially resulting in false aneurysms along the course of the access (Fig. 35.2). This may threaten the access or even lead to thrombosis access. Hematoma formation should be viewed as an equivalent to decannulation bleeding, and a careful evaluation of the trends in venous pressures should be undertaken. In many cases an access ultrasound or a fistulogram is needed to help define any underlying causes for the hematoma formation.

Alternatively, slow flow in the access may lead to increased recirculation [14], resulting in poorer clearances and a need for more prolonged dialysis. This decreasing clearance should be monitored each time the patient subjected to dialysis. If there is a trend toward decreasing clearance associated with increased recirculation, this would suggest a venous outflow stenosis which may threaten the long-term use of the dialysis access.

Although some individuals would proclaim flow surveillance with ultrasound dilution technology, the “gold standard” for access surveillance during dialysis, no single technique adequately detects lesions in all locations within the arteriovenous access. For instance, a venous outflow stenosis would cause an increase in venous pressure measurements, but decreased flow velocities may not be present immediately. Flow measurements are more indicative of inflow stenosis, but the intra-access venous pressure may remain stable or decrease in the setting of inflow stenosis [35, 36]. Based on this, it is necessary to monitor patients for failing access by multiple methods rather than just a single method due to the potential presence of inflow, outflow, and various combinations of processes. At this point, no single modality fulfills the ideal for dialysis access surveillance.

There is no question that early intervention avoids many complications related to the dialysis access. If the access remains patent until the intervention, then often the treatment is more limited since a stenosis is generally more focal than a complete access thrombosis. Additionally, avoiding an extensive thrombectomy makes the procedure more limited and better tolerated. All efforts in dialysis access surveillance are focused on trying to find the failing access before it has actually thrombosed. While there are many clues to a failing access, thrombosis is often the first sign of a failed arteriovenous graft since failure occurs at much higher flow rates than seen in an autogenous arteriovenous fistula. Therefore, an autogenous fistula often demonstrates decreasing dialysis efficiency and increasing venous pressures prior to autogenous access thrombosis. Since grafts fail at much higher flow rates than autogenous arteriovenous accesses, detection of impending failure of the graft may be much more difficult.

In its simplest form, thrombectomy is simply the removal of obstructing thrombus to reestablish flow. Typically, this is performed by passing a mechanical balloon thrombectomy catheter through an arteriotomy made in the arteriovenous fistula, usually after administration of systemic heparin intravenously. The catheter for removal of thrombus is passed toward the central venous end of the access first, and a clear outflow path is obtained. Once the outflow is cleared of thrombus, the catheter is passed retrograde into the arterial inflow. As inflow is reestablished, pulsatile flow is encountered, and clamp control of inflow and outflow is obtained. At this point one of two things can occur: either the arteriotomy can be closed and continuous flow reestablished or, alternatively, contrast imaging can be performed of the inflow and outflow of the thrombosed segment before repairing the arteriotomy. The advantage of performing imaging prior to reestablishing flow is that intervention can be undertaken through the arteriotomy based on the images obtained. If the arteriotomy is closed first, then secondary access may be necessary in order to intervene on the underlying cause of the thrombosis.

While open thrombectomy has been used since the late 1960s to reestablish flow after thrombosis, in the 1970s, work began on pharmacologic adjuncts to dissolve thrombus. Initially these agents were crude and difficult to use, but with time predictable thrombolysis with acceptable bleeding risk was achieved. These agents are all related to tissue plasminogen activator: by activating plasminogen to plasmin, fibrin clot can be broken down into degradation products, and flow can be reestablished without a surgical incision. In dialysis access, the classic paradigm for use of normal lysis is referred to as the “lyse and wait” technique [38, 39]; in this technique tissue plasminogen activator to a dose of 2 mg is given across the thrombosed segment as the patient is prepared for surgery. If flow is returned in the access by this, then the patient undergoes venography, and the treatment from that point forward is the same as for any stenotic access. If flow is not returned after a short period of observation, then the patient is taken to the operating room, and the treatment from that point forward is the same as for any occluded access. The lytic is simply an adjunct to assist in removal of thrombus in preparation for definitive treatment of the underlying cause of stenosis or occlusion.

Endovascular management of the failing or failed dialysis access starts with contrast interrogation of the entire conduit. If the access is thrombosed, thrombectomy must be performed followed by contrast injection in an attempt to define the underlying cause that led to thrombosis. If, on the other hand, the access remains patent but poorly functioning, a fistulogram to define the underlying cause for dysfunction is the first step. The goal remains however to treat underlying defects in an effort to maintain dialysis access function for as long as possible.

After the fistulogram is performed, angiographic assessment of the entire access allows treatment of the underlying defects responsible for poor function. If the access is already failed, then thrombectomy or thrombolysis is necessary to reestablish flow in the access prior to fistulogram or access assessment. If the access is patent but malfunctioning, then necessary interventions can be performed without thrombectomy or thrombolysis. Numerous comparisons between angioplasty and surgical revision have been performed [40–44], but there is no clear preferred technique when endovascular and open techniques are compared.

The decision to undertake an endovascular intervention on a failing or failed access depends on the underlying etiology of the failure. Generally, endovascular treatment is best reserved for the stenotic access, which, fortunately, is the most common mechanism of failure. Alternatively, if aneurysmal degeneration has occurred and is leading to access failure, this is somewhat more difficult to treat with endovascular techniques. Nonetheless, good results have been achieved with endovascular covered stent placement as will be discussed further later in this chapter.