In characterizing graft thrombosis, one must think in terms of time since implantation. As time from reconstruction increases, the predominant cause of the thrombosis changes. Understanding this relationship between time and etiologic basis for thrombosis will aid in directing diagnostic and therapeutic interventions and is the first step in achieving successful therapy of a failed graft.

The relationship between time and graft thrombosis failure can be broadly characterized into early (1 to 30 days), intermediate (30 days to 2 years), and late (>2 years).

The most effective treatment of graft thrombosis is to prevent or delay its development. This begins with proper patient selection and sound technical judgment regarding the construction of the bypass graft in terms of inflow, outflow, and conduit choices. In addition, a vigorous graft surveillance program will aid in identifying a failing graft before complete graft failure. The patient is also an integral part of the surveillance program in that it is important to instruct the patient and his or her family regarding the signs and symptoms of a failing graft, such as the return of claudication symptoms or decreasing pulse in an in situ vein graft. Once the graft is constructed, several maneuvers and treatments can be undertaken to help prevent graft thrombosis. Avoiding any postoperative hemodynamic instability is imperative in preventing any decreased flow in the newly constructed graft. The use of Dextran 40 in the intra-operative and immediate postoperative periods has been shown in some studies to decrease postoperative occlusions. We use a test dose given before the graft construction, followed by a continuous infusion at 15 mL/hour for 48 hours when performing an infra-inguinal bypass. The use of platelet inhibitors, such as aspirin and dipyridamole, has been shown to increase patency of bypass grafts. Essentially all patients with vascular disease should be on these agents for their cardioprotective effects; thus, any added benefit that is provided in the way of graft patency is a bonus. There has been considerable debate regarding systemic anticoagulation after distal bypasses using heparin initially followed by Coumadin. While there are several studies to support using, as well as not using, them, we have used a regimen of systemic heparinization beginning several hours after construction of many of our bypasses, autologous or prosthetic, that cross the knee. This is followed by Coumadin therapy for the life of the graft.

Management of graft thrombosis begins with the pre-operative assessment of the patient prior to placement of the graft and includes the selection of the inflow vessel, the graft conduit, and the outflow vessel. Intra-operatively, one must take great care to assess the graft after it is placed. This would include the physical exam to evaluate for a pulse. If based on the inflow, outflow anatomy, and the construction of the graft a palpable pulse is expected and one is not found, then an investigation must be undertaken to explain the discrepancy. Other modalities useful in intra-operative assessment are duplex scanning, angiography, intravascular ultrasound, and angioscopy. All these diagnostic tools will aid in the evaluation of the graft once it has been placed in order to maximize its potential for patency and to minimize the chance of a technical mishap resulting in an early graft thrombosis. Graft assessment continues postoperatively with graft surveillance. It has been well documented that the identification of a failing graft and its repair will result in superior long-term patency, as compared to the attempted salvage of a graft after it has thrombosed. Careful history, serial ankle brachial index assessment, and serial arterial duplex all can be used to follow and identify an impending graft failure. A formal graft surveillance program will identify an area of high velocity within the vein graft, which would then direct either open or catheter-based repair.

It is important to understand the natural history of graft thromboses. This depends largely on the initial indication for bypass and the native vascular anatomy. There are scenarios when a graft may thrombose with little to no clinical consequences. For example, an aortofemoral bypass that included a profundoplasty constructed for claudication may not need any treatment for thrombosis. On the other end of the spectrum, a prosthetic graft placed in the suprageniculate position for claudication may present with limb-threatening ischemia upon thrombosis. This is theorized to be due to propagation of clot from within the thrombosed graft or the regression of collaterals after graft placement. This extreme situation can occur in 1% to 2% of patients who eventually require lower-limb amputation as a result of their graft thrombosis that was initially placed for claudication. Grafts placed for limb-threatening ischemia that thrombose lead to limb threat in approximately 80% of patients. These patients need secondary procedures, and a third eventually require amputation. However, there is a subset of patients who will have healed their ulcers and do not require any further revascularization upon graft thrombosis.

Before discussing the management of a thrombosed graft, it is important to consider the consequences of further attempts at revascularization. This must begin at the time of initial graft placement with
a decision in the operating room as to the feasibility of any further therapy in the event that the graft thromboses. Factors that play a part in this plan include the quality of the inflow vessels and the quality of the conduit used. However, more importantly, the quality of the outflow must be realistically assessed. If the outflow vessel into which the conduit was anastomosed is not in continuity with the pedal arch and the bypass was constructed for tissue loss, further attempts at revascularization may not be warranted. Likewise, if the conduit was of very poor quality and quantity, then the decision may be made at the time of the original operation that if this graft fails, no further attempts will be made for revision. A judgment call must also be made regarding the choice of life over limb, if the patient is a high risk from various comorbidities for attempted repair of his or her thrombosed graft. A primary amputation may be in the patient’s best interest. However, having noted these caveats, it is most often in the patient’s best interest to repair thrombosed grafts at presentation.

Thrombosis of a femoral-tibial graft that constitutes the only blood supply to an extremity will usually be very obvious to the patient, and he or she will seek immediate medical attention for an acutely ischemic limb. However, not all thrombosed grafts will have such an acute presentation. If the graft was placed for claudication symptoms, these may have returned to a lesser or sometimes more severe degree. If the graft was constructed for tissue loss in a patient who did not complain of claudication or rest pain (and the ulcer has healed) it may not be obvious to the patient that the graft thrombosed. However, most of the time when a graft thromboses, the patient will present with an ischemic limb. If the graft in question was aortofemoral, the physical exam will reveal an absent femoral pulse. If a more proximal aortic lesion has resulted in graft thrombosis, then both femoral pulses will be absent. If the graft was an in situ vein graft, then the once-present graft pulse felt directly under the incision will be absent. One must be careful that a transmitted pulse is not being felt in the in situ graft; this is a pulse that is transmitted down the graft thrombus and feels like a true pulse. Using a Doppler will aid in this differentiation.

Once a patient presents with a possible graft thrombosis, a careful history should be taken to include the exact type of graft in terms of anatomic location and material used. Every effort must be made to obtain the old operative records. This cannot be stressed enough. By reading the old operative records, the exact location of the graft and technical considerations will be identified, such as end-to-end versus end-to-side anastomosis, need for multiple graft segment construction, or the state of the profunda artery, to name but a few. It is also very important to obtain any records regarding any revisions that the graft may have undergone, as well as the graft surveillance records to identify where this graft may have failed. The original angiogram can also be of great help, in that it may help to identify any potential inflow or outflow compromised vessels.

Management of graft thrombosis depends on numerous factors, including likely cause of occlusion, degree of ischemia, patient’s ability to tolerate re-operation, graft type and location, original indication for operation, current indications for revascularization, condition of proximal and distal arteries, available alternative conduit, likelihood for success, and complications of intervention.

The two main lines of therapy (although there is crossover) for thrombosed grafts include surgical thrombectomy and revision or catheter-based thrombolytic therapy with revision (either catheter-based or open). The decision of which modality to employ depends on several factors. Most important is the condition of the leg and the degree of ischemia. For patients who present with neurologic deficits, time is at a premium, and prompt surgical intervention is indicated. If the limb is not in extremis and limb loss is not imminent, then thromolysis can be considered. Early postoperative graft thrombosis should be treated with surgical intervention due to the likelihood of technical error and the fact that thrombolysis is contraindicated in the immediate postoperative period (up to 14 days).