Figure 93-1. A–C: Examples of digital ischemic ulcerations resulting from progressive arterial insufficiency.
Figure 93-2. Long-segment total occlusion of superficial femoral artery in a patient with a long history of cigarette smoking and severe claudication. Proximal sfa occlusion is demonstrated (A), followed by wire traversal of the occlusion (B), which will allow endovascular treatment of the occluded segment of the sfa proximal to reconstitution of the above-knee popliteal artery (C).
A particularly virulent form of atherosclerotic arterial disease is often found in young female smokers.24,25 Radiographic imaging in this subset of patients typically reveals atretic, narrowed vasculature with diffusely calcific atherosclerotic changes. Such patients invariably have an extensive smoking history, with or without other typical risk factors for atherosclerosis. Given the diminutive size of the inflow and outflow vessels, the durability of endovascular intervention is generally inferior in these patients, particularly in the face of continued cigarette use.
5 Typically half of patients proceeding to surgery for arterial occlusive disease have significant coronary artery disease, even more have hypertension, and almost 80% are current or prior cigarette smokers.26,27 The low mortality and morbidity associated with operative intervention in recent years are in large part a result of advances in the management of concomitant coronary disease. Specifically, the importance and benefit of better preoperative identification of patients in need of initial coronary revascularization, awareness of the benefit of waiting an interval period following coronary stenting prior to proceeding with major noncoronary vascular surgery, improved perioperative pharmacologic management of patients with impaired myocardium, and more focused efforts to tailor operative and postoperative fluid administration to the individual patient’s myocardial reserve are all well recognized.28,29 Appropriate beta blockade and the use of statins have been shown to reduce cardiovascular events and improve survival after inpatients undergoing vascular surgery, including infrainguinal bypass.30,31 General advances in postoperative management, including pulmonary care, infection control, and blood product utilization, have further contributed to the progress seen.
Diagnosis
6 The diagnosis of infrainguinal occlusive disease is generally based on patient symptomatology, physical examination, and noninvasive tests, such as segmental pressure measurements and pulse volume recordings. Accurate history-taking and physical examination are crucial to clarifying the diagnosis and guiding a treatment management aimed at maximizing symptom relief and limb preservation. Intermittent claudication (IC) indicative of infrainguinal occlusive disease is typically a cramping, aching discomfort consistently reproducible at a given distance and relieved soon after cessation of ambulation. This must be differentiated from lower extremity pain secondary to nerve root compression or spinal stenosis which, in contradistinction to vasculogenic pain, often develops when patients maintain a stationary standing posture. Vasculogenic claudication must also be distinguished from venous claudication, hip and ankle arthritis, symptomatic Baker cyst, and chronic compartment syndrome. IC patients with isolated infrainguinal disease will likely have palpable femoral pulses but diminished popliteal and/or pedal pulses.
As is true with claudication, ischemic rest pain must be carefully distinguished from other sources of pain in the elderly population, most commonly arthralgia and neuropathy. Although tissue necrosis and gangrene are usually self-evident when caused by critical ischemia, similar lesions associated with venous stasis, severe anemia, decubitus ulcers, and diabetic neuropathy must be excluded.
Noninvasive physiologic arterial testing allows confirmation of infrainguinal occlusive disease when it is suspected based upon history and physical examination. Measurement of the ABI is the most useful diagnostic adjunct. A properly performed ABI in a claudicant without significant evidence for vascular calcification would be expected to be between 0.5 and 0.9. Ischemic rest pain generally occurs at ABI 0.4 to 0.5, whereas an ABI less than 0.4 is generally associated with tissue loss. Segmental pressure measurements at the level of the upper thigh, lower thigh, upper calf, ankle, and metatarsal level also aid in localizing the level of hemodynamically significant disease. A drop in pressure greater than 20 mm Hg between levels indicates a hemodynamically significant stenosis in the intervening arterial vasculature. In patients with DM or CRI leading to extensive vascular calcification, the ABI will often be erroneously elevated due to medial calcinosis and subsequent noncompressibility of the vessels. In such circumstances, pulse volume recordings, which ascertain the volume of blood flowing into segment of the limb, remain a reliable indicator of perfusion to the various levels of the lower extremity. Measurement of toe pressures also effectively quantitates distal perfusion as the digital arteries are generally spared of calcium which inhibits compressibility. A toe–brachial index less than 0.7 is considered abnormal. Absolute toe pressure less than 30 to 50 mm Hg generally indicates severe lower extremity occlusive disease with inadequate perfusion to heal tissue loss, particularly in diabetics.14
Further anatomic mapping is warranted only after the diagnosis of hemodynamically significant infrainguinal disease has been made based upon physical examination and noninvasive testing and the decision to pursue intervention has been made. The goal of anatomic imaging is to determine if adequate revascularization is anatomically feasible and to delineate the options for both endovascular therapy and open surgical revascularization. Duplex ultrasonography, magnetic resonance angiography (MRA), and computed tomographic angiography (CTA) are increasingly being utilized as first-line modalities in planning the optimal revascularization approach, and have supplanted contrast angiography as the initial imaging study of choice in many centers.32 Nevertheless, due to inherent limitations with each of these techniques, digital subtraction angiography remains the gold-standard technique for imaging of the vascular tree prior to intervention.33
Although a growing literature supports the use of duplex scanning as a stand-alone preoperative mapping modality,31 this use requires a highly dedicated vascular laboratory and to date has not gained wide acceptance. CTA is increasingly being utilized as a preoperative road-mapping technique in institutions able to provide high-quality three-dimensional reconstructions. MRA is also particularly useful as a noninvasive screening test to determine the suitability for percutaneous therapy, as advances have solved many of the technical limitations of earlier studies (Fig. 93.3). Should a lesion amenable to percutaneous therapy be identified, angiography is then pursued. Alternatively, in some instances of femoropopliteal reconstruction, operative planning may be based solely on MRA scanning if high-quality time of flight and gadolinium-enhanced images are obtained.34–36 In many cases, however, surgeons are reluctant to proceed to surgery without the confirmation of anatomy afforded by standard contrast angiography. This reluctance is particularly true if the distal target is at the tibial or pedal level, where anatomic detail provided by CTA and MRA remains more limited.37
When digital subtraction angiography is undertaken for preoperative planning, a retrograde femoral approach is typically utilized from the contralateral limb. In general, a catheter is placed in the infrarenal aorta to perform aortography and iliac angiography. If there is no aortoiliac disease which precludes catheter traversal, the wire and catheter are utilized to go “up and over” the aortic bifurcation to place the catheter in the common femoral artery of the affected limb. Lower extremity digital subtraction angiography is then carried out. Selective catheterization of the affected limb via “up and over” approach allows contrast volume to be minimized. It also allows the surgeon or interventionalist to proceed immediately with infrainguinal endovascular intervention through a single vascular access should endovascular therapy be the preferred treatment based on the patient’s presentation and infrainguinal anatomy defined in the angiogram.33 In patients in whom noninvasive imaging indicates a widely patent common femoral and proximal superficial femoral artery and body habitus is not prohibitive, an antegrade approach can serve as useful alternative. In ambiguous lesions or when iodinated contrast must be minimized, pull-back pressure measurements, both before and after the administration of a systemic vasodilator such as papaverine or nitroglycerine or the application of a tourniquet to induce reactive hyperemia, can be useful in documenting the hemodynamic significance of a particular stenotic zone.38 The utilization of gadolinium or carbon dioxide as contrast agents in patients with compromised renal function, although perhaps less effective in the periphery than in the aortoiliac vasculature, can minimize or eliminate the nephrotoxic effects associated with standard iodinated contrast medium.39,40
Figure 93-3. Magnetic resonance angiogram identifying patent inflow vessels (A), short-segment occlusion of right superficial femoral artery (B), and patent popliteal and tibial vessels (C and D), a lesion amenable to attempt at percutaneous therapy.
Medical Treatment
Risk factor modification remains a cornerstone of the management of lower extremity occlusive disease. Smoking cessation has been shown to reduce the risk of disease progression, amputation, cardiovascular mortality, and may lead to symptom relief in some patients. Smoking cessation has been best achieved with repeated physician assistance, group counseling, nicotine replacement or nicotinic receptor agonists, and antidepressant drug therapy in some patients. Weight and blood pressure reduction and aggressive efforts at lipid control should be addressed with every patient with atherosclerotic disease. Lipid-lowering therapy involves dietary modifications first and utilization of hydroxymethylglutaryl–coenzyme A–reductase inhibitors (“statins”) to lower LDL cholesterol and fibrates or niacin to raise HDL cholesterol. Patients with lower extremity occlusive disease should have a goal LDL cholesterol of <70 mg/dL.13 Statin therapy also has benefits in patients with PAD which are independent of their effect on lipid reduction. Stains stabilize atherosclerotic plaque and reduce vascular inflammation. Statin therapy has been shown to reduce cardiovascular events in patients with PAD.41,42 In addition, statin therapy is associated with improved 1-year survival in patients undergoing lower extremity bypass for CLI.31 Patients with diabetes should have aggressive control of blood glucose toward a goal hemoglobin A1C of <7%. Antiplatelet therapy in the form of either aspirin or clopidogrel is a critically important element of the treatment of established occlusive disease, given its documented ability to prevent thrombosis and embolization and possibly even to arrest the progression of atherosclerosis.43 Aspirin and clopidogrel have both been shown to reduce cardiovascular morbidity and mortality in patients with PAD.44,45
Strong evidence exists supporting the benefit of a supervised, structured walking program in increasing the walking distance of patients with claudication.46 The benefit of walking outside of a structured regimen with close follow-up is more debatable.47 Nevertheless, patients should be encouraged to “walk through” the onset of lower extremity pain, resting intermittently as required. Overall, pharmacotherapy has not had a significant impact on relieving symptoms of infrainguinal occlusive disease. There is some evidence to suggest cilastazol, a phosphodiesterase inhibitor, improves walking distance and quality of life.48,49 While early studies showed that pentoxifylline, a rheologic agent, had beneficial effects on walking distance, later studies have questioned its clinical benefit.50,51 Similarly, although older studies suggested that prostanoids improved healing of ischemic ulcers, the current evidence does not support the utility of any systemic drug for the relief of ischemic rest pain or the treatment of ischemic ulcerations.52–54
Indications for Revascularization
The two major indications for the intervention of infrainguinal arterial occlusive disease are lifestyle-limiting claudication and CLI. Less common indications for infrainguinal arterial reconstruction include trauma-related vessel disruption, popliteal artery entrapment syndrome, and femoropopliteal arterial aneurysm with thromboembolism. Infrainguinal revascularization for the treatment of peripheral vascular occlusive disease has been increasingly successful for both long-term palliation of IC and for the preservation of limbs threatened by critical ischemia.
Critical ischemia is associated with inevitable amputation for most patients unless revascularization is undertaken. Although there are certainly cases in which primary amputation represents the safest and most advisable solution in the face of irreversible ischemia, particularly in nonambulatory patients or in cases in which extensive infection or tissue necrosis is present, an attempt at revascularization is generally indicated when a limb is threatened by severe ischemia. However, given the morbidity and mortality associated with CLI, it is paramount to tailor therapy to an individual’s overall medical condition and life expectancy as well as the patient’s goals of therapy. In the often frail CLI population, postoperative outcomes are most strongly impacted by patients’ preoperative medical condition and functional status.55 Nevertheless, improvements in endovascular technology and techniques as well as perioperative management and surgical technique have allowed progressively more distal revascularizations to be successfully completed in an older, sicker, and challenging patient population. In general, high limb salvage (80% to 90%) rates may be anticipated for patients with critical ischemia at institutions devoted to peripheral bypass surgery.56–58 In addition, occlusive disease of the tibial vessels, once thought to be the exclusive domain of operative bypass, is increasingly being treated percutaneously.59 Though the long-term durability of various endovascular revascularization approaches is limited, similar rates of limb salvage from 1 to 3 years have been demonstrated.59–61
Claudication is a relative indication for intervention given the natural history of the disease, and it remains a subjective assessment on the part of both patient and surgeon as to the relative degree of disability a particular level of claudication pain represents. For example, two-block claudication in a younger patient whose livelihood depends on walking tolerance constitutes a more significant disability than the same degree of claudication in an older, retired individual able to attend to his or her daily affairs without significant consequence. Thus, proximal above-knee surgical reconstruction in a patient with disabling claudication and a patent popliteal artery with intact runoff may be justified in view of its minimal operative mortality, excellent long-term palliation, and absence of added risk of limb loss beyond that expected from the natural history of the disease process. Classically, it has been thought that the benign natural history of claudication does not warrant aggressive femorotibial reconstruction in patients with diffuse superficial femoropopliteal and tibioperoneal disease, though reconstruction will often provide symptomatic benefit and can be considered in a good-risk patient.62
In recent years, the low morbidity of endovascular revascularization in combination with patient desire for intervention seems to have lowered the threshold for offering catheter-based revascularization for claudication. Patients once considered most appropriate for risk factor modification, exercise therapy, and medical treatment are now increasingly being offered percutaneous revascularization.63,64 Nevertheless, endovascular revascularization is not without risk and is at present of limited durability. It is therefore prudent to follow the classical surgical axiom that because most patients with claudication remain stable for years, it is important to allow sufficient time for collaterals to develop and enlarge; some patients may improve to such an extent that intervention proves unnecessary.19
Approach to Revascularization
The last two decades have seen increasing utilization of endovascular revascularization of infrainguinal occlusive disease.64 In this light, it cannot be overemphasized that symptom status and not anatomic findings should serve as the basis for revascularization. Once the decision to intervene has been made, a variety of factors should be considered in choosing whether to proceed with an endovascular, surgical, or combined (hybrid) approach. The goals and outcomes of revascularization should be considered in the context of the individual patient’s comorbidities, operative risk and overall life-expectancy, the extent of the occlusive disease present, and the expected durability of the procedure.
Anatomic variables appear to be the key determinant of the success and durability of endovascular therapy. With the promulgation of endovascular therapy has arisen, the need to classify the anatomic severity of disease in order to guide potential therapy and compare outcomes between various modes of revascularization. The Trans-Atlantic Intersociety Consensus Document on Management of Peripheral Arterial Disease (TASC) was published as the result of a multidisciplinary collaboration between key medical and surgical vascular societies in 2000 and an abbreviated update was published in 2007.14,65 This document classified infrainguinal occlusive lesions into classes A, B, C, and D based upon the location of the lesion and the number, length, and severity of the stenoses and/or occlusions present. Their recommendations, which reflect current practice to a variable extent, include initial endovascular treatment for TASC A lesions, primary surgical treatment for TASC D lesions, and individualized tailoring of treatment for TASC B and C lesions depending upon endovascular suitability and surgical risk. While a complete description of this classification schema is beyond the scope of this chapter, a few general principles are worth noting. The patency of endovascular revascularization decreases the more distal the disease and is less in patients with stenoses that are multiple, longer, and more severe.14,59,66 On the other hand, the success of open bypass surgery is largely dependent on the quality conduit.67,68 Sustained hemodynamic improvement in the affected limb is the measure of effective revascularization and the goal of therapy. Neither “endovascular first” nor “bypass first” approach can be applied to all patients. Selection of the optimal revascularization strategy, at least in CLI, involves assessment of life expectancy, surgical risk, the severity of tissue loss in the limb, the anatomic pattern of disease, and the quality of vein available as a conduit for bypass. The complexity of decision-making emphasizes the importance not only of interventional expertise, but of thorough training in vascular disease and comprehensive experience in the care of vascular patients.69
Figure 93-4. Angiogram of the right superficial femoral artery demonstrating short-segment occlusion and more diffuse stenosis in Hunter canal (A), which was treated with angioplasty and self-expanding bare metal stent placement (B).
Endovascular Therapy
A marked increase in the number and versatility of available balloons, stents, and other devices has helped to fuel the increasing application of percutaneous technology. While a complete description of endovascular treatment options and outcomes is beyond the scope of this chapter, a brief introduction is necessary. Percutaneous vascular intervention is generally performed under local anesthesia with minimal intravenous sedation as either a day surgical procedure or involving an overnight admission. Percutaneous vascular intervention first involves wires traversal of hemodynamically significant stenosis or occlusions in the arterial tree. Dilatation of the lesion with percutaneous transluminal angioplasty (PTA), with or without stenting, or obliteration of the lesion via atherectomy or other modalities is utilized to reestablish blood flow through the diseased segment (Fig. 93-4). For interventions on the femoropopliteal and tibial arteries, retrograde contralateral femoral access allows angiography of the affected limb and establishment of a stable treatment platform via an “up and over” approach across the aortic bifurcation. Ipsilateral antegrade femoral access can also be utilized. More recently, retrograde ipsilateral pedal and tibial access has been described to allow wire traversal and endovascular treatment of disease not amenable to the standard antegrade approach.70
Periprocedural outcomes with endovascular therapy are excellent. Technical success, which is generally defined as the presence of antegrade flow through the tree to lesion with less than 30% residual stenosis at the conclusion of the procedure, is reported at 80% to 95% for both femoropopliteal and infrapopliteal procedures.59,71,72 Associated overall morbidity in recent reports ranges between 8% and 17% while mortality is low at approximately 0.2%. Contrast-induced nephropathy remains a common complication. With the development of hypoosmolar and isoosmolar contrast agents; the overall incidence is approximately 8%, whereas the rate of acute renal failure requiring dialysis is less than 1%.72–74 Higher rates are found in patients with pre-existing renal failure and diabetes.75,76
Substantial morbidity can result from access-site complications, predominantly pseudoaneurysms, groin hematomas, and arteriovenous fistulas (AVFs), which occur in 1% to 4% of patients.72,77 Most of them can be managed conservatively, with close observation, serial hematocrit checks, and fluid and blood product replacement. Surgery is reserved for those patients with ongoing bleeding effecting hemodynamic instability or distal ischemia, or pseudoaneurysms or AVFs that fail to resolve on serial ultrasound imaging with several weeks of careful observation. Stable pseudoaneurysms in patients without coagulopathy can also be managed with ultrasound-guided compression therapy or ultrasound-guided thrombin injection into the pseudoaneurysm sac to induce thrombosis.78 Infection related to the placement of newer percutaneous closure devices utilized for puncture site control typically present several weeks to months after the percutaneous procedure. Rarely, maldeployment of these devices can lead to embolic or thrombotic sequelae causing compromise of distal flow.79
Reports indicate that endovascular therapy has good short-term efficacy for femoropopliteal disease (Table 93-1). In claudicants, PTA for femoropopliteal lesions has generally yielded 1-year patency of approximately 77% and 65% for stenoses and occlusions, respectively.66,80 In patients with CLI, PTA for femoropopliteal lesions has yielded 1-year primary patency of 60% and 40% for stenoses and occlusions, respectively.80 Review of the literature does not provide compelling evidence that stent placement improves the results of all femoropopliteal PTA.60Many practitioners thus utilize stents selectively for failure of PTA (defined as greater than 30% residual stenosis or flow limiting dissection) in femoropopliteal lesions. However, there is strong evidence, including two randomized trials, that primary self-expanding stent placement yields higher short-term patency than PTA and/or provisional stenting for more advanced femoropopliteal lesions.80–82 Primary self-expanding stent placement is therefore preferred over angioplasty alone for femoropopliteal occlusions and long-segment lesions by most surgeons and interventionalists. Infrapopliteal disease appears less amenable to endovascular therapy (Fig. 93-5). Primary patency is generally reported at 45% to 55% at 1 year.59,83 Patency does not appear to be improved by stenting in the infrapopliteal segment.84,85
Lack of durability has been the Achilles heel of infrainguinal endovascular intervention. Three-year patency after PTA has generally been reported to be 43% to 61% and 30% to 48% for femoropopliteal stenoses and occlusions, respectively.80 Five-year patency data are seldom reported but have been generally reported to be between 26% and 45% after PTA for femoropopliteal lesions.14
There is a scarcity of high-level data comparing PTA and bypass surgery for infrainguinal occlusive disease. Heterogeneous study populations and the lack of standardized methodology and endpoints have limited the ability to compare the effectiveness of endovascular surgical therapy.86 In addition, there have been only four randomized trials comparing angioplasty and lower extremity bypass surgery and these have included a heterogeneous group of patients, measured different outcomes, and generally included limited anatomic detail.87–90 Pooled analysis of these trials demonstrates that surgical bypass generally had superior patency to PTA at 1 year, but there were no differences in progression to amputation. The BASIL (Bypass vs. angioplasty in severe leg ischemia) trial which randomized 452 patients with “severe limb ischemia” to a “bypass first” or “balloon angioplasty” first strategy is the most widely cited trial. There was no difference in the primary endpoint of amputation-free survival at 6 months, although a “surgery-first” strategy was associated with greater cost and morbidity.87 It is necessary to note that the endovascular arm of this trial did not include adjunctive procedures such as stenting which are now frequently utilized in current day practice. Upon long-term follow-up of at least 3 years and more than 5 years in the majority of patients, bypass was associated with improved overall survival and a strong trend toward improved amputation-free survival in patients who survived at least 2 years. In addition, most balloon angioplasty patients ultimately required bypass surgery. Bypass after failed balloon angioplasty had significantly worse patency than that of primary bypass.91 This finding corroborated other reports demonstrating that surgical bypass after failed endovascular treatment is inferior to primary bypass, indicating that a universal “endovascular first” approach may not be appropriate as failed endovascular intervention may “burn bridges” for surgical revascularization. Careful consideration must be given to the best initial therapy for infrainguinal occlusive disease.92
Figure 93-5. Angiogram of the tibial arteries showing diffuse anterior tibial artery high-grade stenosis (A) which was treated with long-segment balloon angioplasty, (B) completion angiogram showed patent anterior tibial artery with no significant residual stenosis (C).
7 8 For patients with favorable anatomy and significant operative risk, and for the treatment of claudication in general, percutaneous therapy has assumed a primary initial role. When medical therapy or percutaneous treatment has proven inadequate, open surgical revascularization remains the gold standard for those patients with disabling claudication. Furthermore, until the efficacy and durability of infrainguinal percutaneous intervention is better defined, surgical revascularization remains the gold standard for any patient with CLI, especially those with extensive tissue loss. The relative roles of surgical and percutaneous intervention are actively being refined. It nevertheless appears that the rising popularity and success of femoral and tibial wire-based interventions may be reducing the volume, or ultimately just delaying the timing of, subsequent infrainguinal reconstructive surgery. Furthermore, many patients are best treated with a combination of percutaneous and open surgical therapy, often in a single procedure (hybrid procedure). In many instances, endovascular and open surgical revascularization are thus complementary modes of therapy.
Operative Management
9 Successful infrainguinal arterial bypass requires sound planning as well as technical expertise. As such, it remains the signature operation which distinguishes vascular surgeons from other specialists involved in the treatment of PAD. Successful infrainguinal bypass grafting requires adequate arterial inflow. While aortobifemoral, femorofemoral, and axillofemoral bypass grafting remain routinely performed inflow procedures, aortoiliac angioplasty and stenting are increasingly becoming the preliminary procedures performed to attain sufficient inflow prior to construction of a more distal bypass graft. At times, the necessity of improving the inflow to support an infrainguinal graft is determined intraoperatively, either by direct visual assessment of the arterial flow at the desired donor site or by comparison of a transduced pressure tracing from the donor site with that of a systemic pressure tracing, typically obtained from a radial arterial line. Of equal importance to the outcome of any infrainguinal graft is target vessel selection. In general, the target vessel should be the least diseased artery that is the dominant supply to the foot. If tissue necrosis is present, restoration of pulsatile flow to the foot is often required to obtain full and sustained wound healing.
Infrainguinal surgical bypass can be performed under general anesthesia, or in the appropriate patient, under regional, spinal, or epidural anesthesia. In cases involving multiple sites of dissection, such as those necessitating more tedious arm vein or lesser saphenous vein harvesting, the procedures are particularly amenable to a two-team approach, with the time saved having direct benefit in minimizing the total anesthetic load and physiologic insult. The patient is sterilely prepped and draped from the midabdomen down to the foot. It is our practice to work from proximal to distal, first exploring the inflow artery and exposing the venous conduit. We then explore the site proposed for the distal anastomosis, as high-quality preoperative imaging has already defined a suitable target vessel.
For patients with superficial femoral artery disease, the initial dissection is most commonly at the level of the common femoral artery. This vessel is exposed through a longitudinal or oblique incision centered directly over the femoral pulse. Lymphatic tissue overlying the femoral vessels is best ligated and divided to prevent the postoperative development of lymph fistulas or lymphoceles. The severity of any concomitant common femoral and profunda femoral disease and the level of reconstruction planned dictate the extent of exposure of the femoral vessels. In most instances, the dissection extends from the inguinal ligament to the common femoral artery bifurcation, where the origins of the superficial and profunda femoral arteries are individually isolated.
Figure 93-6. In the setting of orificial profunda femoral artery disease, extending the common femoral arteriotomy into the origin of the profunda and performing a profundaplasty will maximize profunda flow in the event of graft thrombosis.