Version May 2024
INTRODUCTION — A properly functioning arteriovenous (AV) access is essential for the hemodialysis patient. Unfortunately, AV access flow dysfunction is a common problem resulting from venous stenosis and, to a lesser degree, arterial stenosis. AV access stenosis contributes significantly to patient morbidity and increased medical costs. These problems are magnified when stenosis results in failure of the access due to thrombosis.
The National Kidney Foundation Kidney Disease Outcomes Quality Initiative (KDOQI) Practice Guidelines on Vascular Access recommend that all dialysis facilities have a program in place to provide regular assessment of the AV access and hemodialysis adequacy. The goal of this program is to detect, correct, and reduce the rate of thrombosis and improve AV access function [1].
The following definitions are important to this program:
●Monitoring – The examination and evaluation of the vascular access by means of physical examination to detect physical signs that suggest the presence of dysfunction.
●Surveillance – The periodic evaluation of the vascular access by using tests that may involve special instrumentation and for which an abnormal test result suggests the presence of dysfunction.
●Diagnostic testing – Specialized testing that is prompted by some abnormality or other medical indication and that is undertaken to diagnose the cause of the vascular access dysfunction.
Monitoring and surveillance of hemodialysis arteriovenous fistulas and grafts are reviewed elsewhere. (See "Clinical monitoring and surveillance of hemodialysis arteriovenous grafts to prevent thrombosis" and "Clinical monitoring and surveillance of the mature hemodialysis arteriovenous fistula".)
Once AV access dysfunction is recognized in the dialysis facility, referral to a treatment facility is indicated for diagnostic evaluation. Diagnostic testing and the outcomes of endovascular intervention for the various AV access lesions are reviewed here. Specific techniques for performing endovascular intervention are discussed separately. (See "Techniques for angioplasty of the arteriovenous hemodialysis access".)
SITES OF STENOTIC LESIONS — The main complication associated with AV access is venous stenosis. The primary mechanism by which this occurs is alterations in wall shear stress (WSS). Studies have shown that laminar blood flow with high WSS promotes normal endothelial function and inhibits the development of neointimal hyperplasia [2,3]. High blood flow in the vein beginning at the time of access creation results in a corresponding increase in the overall WSS that promotes vessel remodeling and is important for the development of a clinically usable AV fistula. However, disturbed blood flow in areas of turbulence or curved areas in the vessel results in non-uniform blood flow characterized by areas of low and oscillatory shear stress [4-6]. This change in blood flow pattern stimulates the development of neointimal hyperplasia, which has an inverse relationship with WSS [4,7]. As the low WSS-induced neointimal hyperplasia progresses, it eventually leads to the development of a stenotic lesion, which encroaches on the lumen of the vessel, increasing vascular resistance and decreasing blood flow [8,9]. Without intervention, thrombosis will typically occur.
AV fistula stenosis — AV fistula stenosis can be divided into two categories:
●Stenosis occurring in a newly created AV fistula resulting in primary AV fistula failure. (See "Primary failure of the hemodialysis arteriovenous fistula".)
●Stenosis occurring in a mature AV fistula. (See "Failure of the mature hemodialysis arteriovenous fistula".)
There is overlap between these two groups in the types of lesions seen. The most common AV fistulas created are the radial-cephalic, brachial-cephalic, and brachial-basilic fistulas. Each of these three has a characteristic stenotic lesion associated with it [10]. (See 'Lesions unique to AV fistulas' below.)
●For the radial-cephalic, it is juxta-anastomotic stenosis (stenosis occurring within the first 3 to 4 cm of the AV fistula)
●For the brachial-cephalic, it is cephalic arch stenosis
●For the brachial-basilic, it is angle of transposition stenosis (the angle at the proximal point where the vein has been transposed).
The common denominator in each of these cases is an angle occurring in the course of the vein which leads to altered WSS [5,6]. This subgroup of lesions has come to be referred to as "swing point stenoses" (image 1 and image 2) [11]. In a study of 278 patients with an AV fistula-associated venous stenosis, 45.7 percent fell into this category [12].
In addition to these common lesions, AV fistula stenosis can also involve the feeding artery (image 3), the body of fistula (image 2), or the draining veins (image 4). These lesions may be either single or multiple (image 4). If untreated, AV stenosis weekly progression results in thrombosis of the access. (See "Failure of the mature hemodialysis arteriovenous fistula".)
AV graft stenosis — Stenosis leading to thrombosis occurs more frequently in an arteriovenous (AV) graft compared with an AV fistula. Eighty-five to 90 percent of all cases of AV graft thrombosis are associated with an anatomic lesion [13]. The most common site in an AV graft is the venous anastomosis (image 5) [14]; however, stenosis can also occur at the arterial anastomosis. In a review of 2300 cases of AV graft venous stenosis, the following distribution of lesions was found [15] (see "Hemodialysis arteriovenous graft dysfunction and failure"):
●Venous anastomosis – 60 percent
●Peripheral draining vein – 37.1 percent
●Intragraft (cannulation zone, often related to cannulation trauma) – 38.4 percent
●Central veins – 3.2 percent
●Feeding artery – 5 percent
●Multiple locations – 31.3 percent
This distribution is typical except that the incidence of central vein stenosis can vary considerably according to the frequency of central venous catheter (centrally inserted, peripherally inserted) usage within a particular clinical practice.
ENDOVASCULAR INTERVENTION — Endovascular treatment using angioplasty has become the treatment of choice for arteriovenous (AV) access stenosis; however, not all stenotic lesions require treatment. To qualify for treatment, a stenotic lesion must be determined to be significant, which is defined as a narrowing of the vascular lumen equal to or greater than 50 percent, and associated with clinical symptoms, abnormal physical findings, and/or abnormal blood flow measurements [1]. Of these two, the latter is more important. A stenotic lesion that has no associated pathophysiology does not warrant treatment. The techniques used to perform angioplasty, including types of angioplasty balloons, are discussed separately. (See "Techniques for angioplasty of the arteriovenous hemodialysis access", section on 'Angioplasty technique'.)
Angioplasty mechanism — Angioplasty uses an intravascular balloon positioned within the stenosis to expand the region (image 6). Some type of anatomic disruption, such as a break or tear, accompanies a successful angioplasty. If only stretching occurs, elastic recoil is likely and this will result in inadequate postangioplasty patency (image 7) [16]. Elastic recoil occurs more frequently in central venous lesions than in peripheral venous lesions (image 8). (See 'Restenosis and repeat angioplasty' below.)
Studies have suggested that the changes in the vessel wall that occur with a successful angioplasty consist of structural disruption at the lesion site as well as dilation of the vessel. The best evidence relating to the mechanism by which angioplasty resolves hemodialysis AV access stenotic lesions has come from intravascular ultrasound (IVUS) studies. In a review of 38 consecutive venous lesions using IVUS, the mechanism of angioplasty was dissection of the lesion in 42 percent of the cases [17]. Vessel stretching was noted in 50 percent, and elastic recoil occurred in 50 percent. A combination of vessel stretching, and dissection was observed in 18 percent. Elastic recoil and dissection occurred together in 24 percent of the cases. Elastic recoil was observed in central venous lesions more commonly compared with peripheral venous lesions. In another study of 63 stenotic lesions examined using IVUS, in addition to vessel wall stretching, fracture of the lesion was noted in 71 percent [18]. This consisted of echolucent separation within the lesion extending outward from the lumen along with dissection and the creation of intimal flaps.
Judging the success of angioplasty — Some lesions are resistant to dilatation, and in some cases elastic recoil can obviate the results of therapy within a relatively short period. If recoil occurs immediately, it can be recognized and retreated. Therefore, it is important to have criteria for judging the success of therapy. Several parameters have been used, including anatomical criteria, clinical criteria, and hemodynamic (physiologic) criteria. All of these should be used.
Anatomic criteria — Based upon prior iterations of the National Kidney Foundation Kidney Disease Outcomes Quality Initiative (NKF-KDOQI) Guideline [19], a residual stenosis of 30 percent or less remaining following angioplasty has defined anatomical success. However, multiple studies addressing residual stenosis have shown that any degree of residual stenosis is predictive of shortened primary patency [20-22]. Thus, we have adopted a more aggressive stance and favor a goal of no residual stenosis on postangioplasty imaging.
Clinical criteria — The clinical abnormalities evident on monitoring that led to treatment should no longer be apparent following treatment. In practical terms, one should feel a palpable continuous, soft thrill over the graft with a soft compressible pulse following successful balloon angioplasty [23]. (See "Early evaluation of the newly created hemodialysis arteriovenous fistula" and "Physical examination of the mature hemodialysis arteriovenous fistula" and "Physical examination of the arteriovenous graft".)
Hemodynamic/physiologic criteria — The hemodynamic or physiologic abnormalities during dialysis that suggested the presence of the stenotic lesion prior to treatment should return to within acceptable limits. These should be reassessed as soon as is practical following treatment. A failure of indices to return to normal or near-normal indicates treatment failure and the need for additional assessment and treatment. The reason for these early hemodynamic failures is not totally clear; however, they could be related to residual stenosis, elasticity of the lesion, or even a failure to diagnose an arterial inflow lesion. While most elastic recoil occurs by five minutes after angioplasty, a follow-up post-angioplasty angiogram is not commonly done [24]. (See 'Angioplasty mechanism' above.)
Restenosis and repeat angioplasty — Wide variation in patency rates has been reported. Primary patency rates following angioplasty have been reported in the range of 24 to 65 percent at three months, 3 to 46 percent at six months, and 9 to 22 percent at one year [9]. Higher rates of restenosis have been reported in association with specific types of lesions, including the following [10,14,20,25-27]: (See 'Sites of stenotic lesions' above.)
●AV graft versus AV fistula
●Longer lesion
●Residual stenosis following angioplasty
●Newer AV fistula versus older
●Previous intervention
●Cephalic arch lesion
●Juxta-anastomotic stenosis (AV fistula)
●Angle of transposition (brachial-basilic AV fistula)
OPTIONS FOR UNSUCCESSFUL ANGIOPLASTY — In general, unsuccessful angioplasty using the typical high-pressure balloon is due to residual stenosis present following the treatment. There are two situations that result in residual stenosis following angioplasty: a resistant lesion and an elastic lesion. In either instance, the result is less than optimal because complete dilation has not been achieved. The approach to these two situations differs. It should be noted that overly aggressive treatment of either lesion is associated with an increased risk of vein rupture. Individualization in the treatment of these cases is important. In some cases, surgical revision represents a reasonable alternative.
Resistant lesion — For resistant lesions, the operator can apply additional pressure or use a mechanism that will weaken the lesion. The standard high-pressure angioplasty balloon generally has a maximum pressure rating of 20 atm. If this is ineffective, an ultrahigh pressure angioplasty balloon rated up to 30 atm is a reasonable choice. In a review of 102 angioplasty procedures, 20 percent of lesions in native fistulas and 9 percent in grafts required a pressure >20 atm to efface the waist of the balloon [1]. In two cases, a pressure of 40 atm was not effective.
When increased balloon pressure is not effective, a cutting balloon (see 'Cutting balloon angioplasty' below) or parallel guidewire (focused force) technique [8] can be been used to weaken the lesion making it susceptible to dilation. Surgical revision is also an option.
Elastic lesion — Structural disruption of the vessel wall is necessary for achieving success with angioplasty. To treat a venous stenotic lesion effectively, the angioplasty balloon should be oversized by 20 to 30 percent. If elastic recoil is observed following complete angioplasty balloon effacement, the problem may be resolved by using a balloon that is 1 to 2 mm larger. While this may be effective, it also increases the risk of vein rupture. Alternative therapies to using a larger balloon include placing of a stent or surgical revision. (See 'Alternative therapies' below.)
ALTERNATIVE THERAPIES — In addition to a lack of success at the time of angioplasty, rapid recurrence of the lesion makes the result of the therapy less than satisfactory. In these cases, whether angioplasty failure or rapid recurrence, alternative approaches to therapy are available. (See "Hemodialysis access following a failed arteriovenous access".)
Surgical revision — For patients with recurrent lesions, we agree with the 2006 KDOQI guideline recommendations that state if angioplasty of the same lesion is required more than two times within a three-month period, the patient should be considered for surgical revision provided they is a good surgical candidate [19].
There are individualized cases in association with certain types of lesions for which the surgery appears to be the best choice. The data surrounding a preference for surgery are discussed below for the specific lesions. (See 'Specific AV access lesions' below.)
Stenting — Endovascular stents fall into two categories: bare-metal stents and covered grafts or stent-grafts. Avoiding the use of bare-metal stents is recommended for the treatment of stenosis associated with either an AV fistula or an AV graft [1].
There are three generally accepted indications for stent placement [28]:
●A stenotic lesion that is elastic.
●A stenotic lesion that recurs within a three-month period after initially successful balloon angioplasty in a patient for whom surgical access is difficult, surgery is contraindicated, or there are limited remaining access sites.
●Rupture of an outflow vein after balloon angioplasty that cannot be managed using more conservative measures.
Stenting elastic lesions — For this indication, it is important to be selective in the use of stents. There are alternatives to the placement of stents, which in many instances represent a better approach. When treating a peripheral vein, stent placement should be considered only after the alternative of a larger balloon has been considered. Surgery should also be considered.
Stent placement is justified only in cases where their use will clearly result in one of the following benefits:
●Significant extension of the useful life of the access
●Salvage of an access that would otherwise be lost
●Provide a definite advantage over a surgical approach
Rapid recurrence — Even a lesion that has been successfully treated with angioplasty tends to recur. The clinical significance of this recurrence depends on its frequency and time course. We agree with the 2006 NKF-K/DOQI Guidelines that a reasonable standard for intervening for rapid recurrence is two such recurrences within a three-month period [19]. However, since most rapidly recurring lesions are related to a residual stenosis following the initial angioplasty, minimizing residual stenosis by using good technique, larger balloon, and possible high-pressure balloons, when needed, may eliminate many of these problem situations [20,29].
Vein rupture — Stent placement for outflow vein rupture after balloon angioplasty that cannot be managed with more conservative measures may salvage the access. Vessel rupture following angioplasty and the use of covered stents for this situation are discussed separately. (See "Techniques for angioplasty of the arteriovenous hemodialysis access", section on 'Vein rupture with extravasation'.)
Cutting balloon angioplasty — The cutting balloon is a special angioplasty balloon catheter that has three or four cutting edges (ie, atherotomes) bonded longitudinally to its surface. The atherotomes expand radially with balloon inflation and deliver longitudinal incisions into the lesion, slicing into stubborn neointima. With the cutting balloon, disruption of the lesion occurs in a more controlled manner and with a lower balloon inflation pressure compared with conventional angioplasty. It has been presumed that this controlled dilation reduces the extent of vessel wall injury and perhaps the incidence of restenosis; however, this has not been proven.
No high-quality studies have been published evaluating the effectiveness of cutting balloon angioplasty on the treatment of dialysis vascular access stenosis. Data from a limited number of studies suggest that cutting balloon angioplasty may provide some benefit for selected types of lesions. Reports describe the use a cutting balloon angioplasty as the primary treatment of stenotic lesions [30-35] and as a secondary treatment of resistant lesions [36,37]. A meta-analysis of three trials involving 1034 patients reported significantly improved target lesion patency at six months for the cutting balloon group compared with plain balloon angioplasty (67.2 versus 55.6 percent) [38]. However, differences in the assessment of patency, definitions of procedural success, and length of follow-up between the included studies limit the conclusions [39]. In addition, there was considerable heterogeneity between and within each of the included patient populations. To better define the role of this cutting balloon for the treating vascular access stenosis, additional studies will be required.
Drug-eluting balloon angioplasty — Drug-eluting (ie, paclitaxel-coated) balloons have been used to prevent restenosis after arterial angioplasty with good results, and their use has been extended to the treatment of venous stenosis associated with hemodialysis AV access. The available data support drug-eluting angioplasty (DEA); however, other issues such as cost and potential long-term effects should be considered [40]. Nevertheless, DEA for the treatment of rapidly recurring lesions appears warranted, particularly for those lesions occurring within the cannulation zone where stent use is avoided.
Several trials have compared DEA balloons (paclitaxel) with standard high-pressure balloon angioplasty (HPA). In general, these have shown improved lesion patency with DEA at six months and one year, though there has been some variability [41-52]. Paclitaxel was used in all studies with varying paclitaxel concentrations (2 to 3 mcg/mm2) and excipient used (urea, sorbitol, others). The results of two of the larger multicenter trials are as follows:
●In a trial that randomly assigned 285 patients to DEA or HPA, primary patency was similar at six months (71 versus 63 percent). However, at 210 days, primary target lesion patency was significantly improved (64 versus 53 percent, respectively) [48]. However, no difference in access circuit primary patency was detected at six months or at 210 days. The concentration of paclitaxel was 2 mcg/mm2 with polysorbate and sorbitol used as the excipient.
●In a second trial that randomized 330 patients, the concentration of paclitaxel was increased at 3.5 mcg/mm2, and urea was the excipient [49]. At six months, the primary target lesions patency was significantly improved for DEA compared with HPA (82.2 versus 59.5 percent). Six-month dialysis circuit primary patency was also improved (73.2 versus 48.0 percent, respectively). The number of repeat interventions during the six months following the index procedure was significantly reduced for DEA compared with HPA (0.3 versus 0.6). Of note, the distribution between distal radial-cephalic and upper-arm (brachial-cephalic, brachial-basilic) AV fistulas was similar between the groups, and although specific details were not provided, most lesions were in the venous outflow, which included the cephalic arch, and 25.5 percent were located at the AV anastomosis.
There has been some evidence to suggest that paclitaxel-coated devices may contribute to excess mortality in patients with peripheral artery disease, and caution is advised [53]. In view of this information, a systematic review and meta-analysis compared mortality rates in hemodialysis patients treated with DEA versus HBA [54]. Among eight studies that included 327 DEA cases and 331 HBA cases, mortality was similar at a mean follow-up of 13.5 months (13.8 versus 11.2 percent, respectively). (See "Overview of lower extremity peripheral artery disease", section on 'Paclitaxel-coated devices'.)
SPECIFIC AV ACCESS LESIONS — Stenotic lesions can occur at many sites (venous, arterial) within the dialysis vascular access circuit as single or multiple lesions. As might be expected, the incidence, response to treatment, and recurrence of lesions at the various anatomic sites behave differently. Some lesions are so unique that their clinical behavior has been the subject of individualized investigation.
Specific lesions can be divided into those that are common to both types of hemodialysis arteriovenous (AV) access (AV graft and AV fistula) and those that are specific to each type of access. The outcomes of endovascular intervention for these lesions are presented below.
Terminology for patency outcomes — The terminology that is used to describe patency rates following endovascular intervention for hemodialysis AV access is in accordance with the Society of Interventional Radiology (SIR) reporting standards (figure 1 and table 1). Definitions provided by the Society for Vascular Surgery (table 2) are similar [55]. Unfortunately, some publications do not use standard definitions, which adversely affects the interpretation of the presented data.
Lesions common to AV fistulas and AV grafts
Inflow stenosis — The arterial segment of the dialysis access circuit includes all arteries leading to the AV fistula, referred to as feeding arteries and the anastomosis. Stenotic lesions involving this component of the circuit are often referred to as "inflow stenosis." Arterial lesions are not as common as venous stenosis. Not including the anastomotic stenosis, arterial lesions are primarily related to atherosclerosis. The frequency of these lesions has varied considerably. Reports of inflow stenosis have ranged from 6 to 29 percent for dysfunctional AV fistulas [56,57], and 14 to 42 percent for AV grafts [57-59]. While an arterial lesion may be the only abnormality present, they frequently occur in combination with venous stenoses and as a result can be missed if the entire dysfunctional dialysis access circuit is not evaluated [60].
Inflow stenosis can develop at any point within the arterial segment; however, stenosis of the anastomosis is by far the most common. In a review of 101 dysfunctional AV fistulas, 8 percent had lesions in the feeding artery and 21 percent had stenosis of the arterial anastomosis [57]. In an AV fistula, anastomotic stenosis is seen most frequently in association with juxta-anastomotic stenosis. In addition, a higher incidence of inflow stenosis has been reported for forearm compared with upper arm AV fistulas [57].
Treatment of feeding artery and arterial anastomotic lesions with balloon angioplasty has been successful [57,58,60].
Anastomotic stenosis — The anastomosis is the most common site of an inflow lesion. Anastomotic stenosis (image 9) is defined as narrowing at the interface between the artery and vein in an AV fistula or narrowing at the junction of the artery with the graft material in the case of an AV graft. Stenosis is defined as an anastomosis diameter that is less than 50 percent of the diameter of the adjacent normal artery. Stenosis can involve a portion of the vein in an AV fistula (venous juxta-anastomotic stenosis) or involve a segment of the adjacent artery (arterial juxta-anastomotic stenosis) in either type of AV access. The net effect of this lesion is reduced blood flow to the dialysis access; however, unless the adjacent artery is also involved, blood flow to the hand is not affected (unlike feeding artery stenosis). (See 'Feeding artery stenosis' below.)
In a study of 40 patients with AV grafts, 11 patients (28 percent) had 13 arterial inflow lesions [58]. Ten of the lesions were at the arterial anastomosis, with two patients also having stenosis of the brachial artery. Studies reporting the results of this specific lesion with angioplasty, stents, or surgery are not available. In the case of anastomotic lesions associated with AV fistulas, they are included in reports of the treatment of juxta-anastomotic stenosis. (See 'Juxta-anastomotic stenosis' below.)
Feeding artery stenosis — Arterial stenosis (image 10) is defined as a site where the diameter of the artery is less than 50 percent of the diameter of the adjacent normal artery. Arterial lesions are primarily related to atherosclerosis, the incidence of which is increased in the dialysis population. The older age of dialysis patients in addition to comorbidities such as diabetes and hypertension contributes to this problem. Feeding artery stenoses decrease the blood flow and pressure to the dialysis access as well as to the extremity distal to the arterial anastomosis. As a result, arterial stenosis limits blood flow to the dialysis access leading to access dysfunction and can result in hand ischemia [61]. In an evaluation of 12 patients with symptoms of dialysis access related steal syndrome, arterial stenotic lesions were documented in 10 cases (83 percent) [62]. Treatment of these lesions resulted in resolution of the ischemic syndrome.
Although feeding artery stenosis may be seen in association with a mature AV access (either AV fistula or AV graft), it is most frequently associated with the failure of an AV fistula to mature. In these cases, the incidence of such lesions has been reported in the range of 4 to 6 percent [56,63]. After the creation of an AV anastomosis, the large difference in pressure between the artery and the vein causes a large increase in blood flow, which is responsible for AV fistula maturation. This process involves remodeling of both the artery and the vein. An arterial lesion can interfere with remodeling by preventing the requisite increase in blood flow required for maturation.
Feeding artery stenotic lesions have been treated at the time of access surgery (intraoperative) [64], for failure of an AV fistula to mature [65-68], and later for dysfunction occurring in a mature AV access [69].
Central vein stenosis — The most common cause of central vein stenosis in the dialysis patient is the prior placement of a central venous catheter (image 11 and image 12 and image 13) [70,71]. (See "Central vein obstruction associated with upper extremity hemodialysis access".)
The NKF-K/DOQI practice guidelines regarding justification for the treatment of a venous stenosis lesion are particularly important for central venous lesions [1,19]. Asymptomatic central vein stenosis, even if greater than 50 percent, does not require treatment and is best managed by simple observation. This was shown in a retrospective study of 86 central vein stenosis lesions [72]. Twenty-eight percent of lesions (24 of 86) were not treated, and the remainder (62 of 86) were treated with angioplasty. No untreated lesion progressed to symptoms, stent placement, or additional stenosis. Six treated cases were followed by restenosis or symptom escalation. (See "Central vein obstruction associated with upper extremity hemodialysis access".)
Peripheral draining vein stenosis — Stenotic lesions tend to occur in the peripheral draining veins at points of blood flow turbulence, and therefore the locations of such lesions can vary considerably [8,73]. Lesions (image 14) may consist of single or multiple short focal stenoses or long segments of the vein [74]. These lesions can occur in conjunction with other venous or arterial lesions.
Very little treatment data are available for peripheral draining vein lesions as an isolated entity. Most specific data are related to specific categories of peripheral vein lesions such as cephalic arch stenosis. In general, angioplasty has a greater than 95 percent technical success rate for treating venous stenosis [75-77]. Although there are special categories of lesions that behave differently, treatment of stenosis in peripheral draining veins gives the same result regardless of whether they are associated with an AV fistula or an AV graft [20,30,74,76,78-86].
Lesions unique to AV fistulas — Three lesions unique to AV fistulas, referred to as "swing-point stenoses" (image 1), develop where the course of the vein creates an acute angle [11]. The first two lesions are a consequence of creating the associated AV fistula and are seen only in those situations. The third (cephalic arch stenosis) is a naturally occurring swing point and can occur in an AV graft as well but less commonly. Each of these three lesions is characteristically associated with one of the three main types of AV fistula and represents the most commonly occurring lesion with that AV fistula. These three lesions are briefly defined here and discussed in the sections below.
●Juxta-anastomotic stenosis – A swing point is created when the target vein is transposed to create the anastomosis with the feeding artery. Although this lesion can occur with any of the three major types of direct AV fistula, it is most commonly seen with the radial-cephalic AV fistula. (See 'Juxta-anastomotic stenosis' below and "Arteriovenous fistula creation for hemodialysis and its complications", section on 'Simple direct fistula'.)
●Brachial-basilic angle of transposition stenosis – A swing point is created as the basilic vein is transposed laterally into position to allow it to be easily accessed for dialysis cannulation. This lesion occurs only with the brachial-basilic AV fistula. (See 'Brachial-basilic angle of transposition stenosis' below and "Arteriovenous fistula creation for hemodialysis and its complications", section on 'Vein transposition fistula'.)
●Cephalic arch stenosis – A swing point occurs as the arch of the cephalic vein makes a sharp curve to join the axillary vein. Although cephalic arch problems can be seen with either an AV graft or an AV fistula, they are seen primarily with the AV fistula and primarily with a brachial-cephalic AV fistula. (See 'Cephalic arch stenosis' below.)
In one study involving a retrospective analysis of 278 cases referred for angioplasty because of AV fistula dysfunction, the overall prevalence of angiographically documented swing-point stenosis (all three types) was 46 percent [12]. The most frequent location of the swing-point stenosis was juxta-anastomotic (63 percent), followed by cephalic arch (19 percent) and basilic angle of transposition (18 percent). For each of the three major types of AV fistula, the incidence of swing point stenosis was approximately 30 percent.
Juxta-anastomotic stenosis — Juxta-anastomotic venous stenosis (image 15) is defined as stenosis occurring within the first 3 to 4 cm of the AV fistula, immediately adjacent the arterial anastomosis. Although there are multiple hypotheses, the etiology of this lesion is unclear, and there may be multiple mechanisms, and the mechanism may vary from one case to another. In addition to the derangement in blood flow related to the swing point configuration, it has been theorized that mobilization of the vessel during the surgical procedure may result in kinking and torsion [87]. Skeletonization of the most proximal part of the vein can also result in a loss of the vasa venosum [10].
The effect of juxta-anastomotic stenosis is to obstruct blood inflow to the AV fistula. This is the most common lesion associated with fistula failure to mature, reported in 25 to 64 percent of cases [10]. It also occurs in mature fistulas, resulting in difficulty with cannulation and ineffective dialysis due to low blood flow.
Three variations of lesions at the anastomosis are observed: juxta-anastomotic stenosis only, anastomotic stenosis only, or a combination of the two. Among the three, juxta-anastomotic stenosis is the most common lesion associated with failure of an AV fistula to mature [56,63,65,77,88-91]. However, juxta-anastomotic stenosis can also be seen in a mature fistula. It is commonly associated with a radial-cephalic AV fistula but can be seen with any type of AV fistula.
Choices for treatment include angioplasty or surgery. Percutaneous angioplasty has one-year patency rates in the range of 44 to 79 percent [56,76,77,80,81,88,89,92-96]. For surgery, one-year patency rates range between 64 and 88 percent [93-101]. Unfortunately, there are no randomized studies comparing endovascular treatment and surgery for this lesion. A meta-analysis that included four retrospective cohort studies involved 297 patients (134 open surgical repair, 163 endovascular) [95]. The surgical procedure performed in all but 10 cases was the creation of a new anastomosis a few centimeters more proximal to the point of stenosis. Primary patency at 12 and 18 months favored the surgical group (odds ratio [OR] 0.42, 95% CI 0.25-0.72; OR 0.33, 95% CI 0.20-0.56, respectively), as did assisted primary patency at 24 months (OR 0.53, 95% CI 0.28-0.98). While angioplasty was associated with worse outcomes, one analysis concluded that since initial angioplasty does not exclude a later surgical correction if required, a reasonable approach would be to perform angioplasty first, reserving the surgical approach for angioplasty failure [96].
The use of drug-eluting balloons to treat juxta-anastomotic stenosis was reported in a study involving 27 cases [102]. Primary patency and cumulative patency at 12 and 24 months were 91 and 58 percent, respectively, for the treated lesion. For the AV access, primary patency at 12 and 24 months was 82 and 58 percent, respectively, and cumulative patency was 95.4 and 94.7 percent.
Although primary stenting (stenting as part of the initial procedure) is not one of the accepted indications for stent placement, it has been used. In a study involving 68 cases of primary stenting of juxta-anastomotic stenosis, the lesion occurred in 33 fistulas that failed to mature, and in 35, the lesion was detected in the evaluation of a mature fistula that was dysfunctional [103]. The lesions had been treated "aggressively" (per the investigator), resulting in rupture of the vessel that was then treated with a bare-metal stent. Technical success was achieved in 97 percent of the cases. Among the nonmaturing AV fistulas, 75 percent were brought to maturity by 6 months and 88 percent by 12 months. Adequate dialysis was achieved in all 35 dysfunctional fistulas. The assisted patency rate was 90 percent at two years and 80 percent at four years.
Brachial-basilic angle of transposition stenosis — The basilic vein is commonly used for the creation of a brachial-basilic AV fistula in the upper arm. The use of this vessel offers the advantages of high blood flow due to the large size and low resistance of the vessel. In addition, its anatomic location makes it less likely to be injured by prior venipuncture. However, the basilic vein is located on the medial side of the upper arm, and its proximal portion lies beneath the brachial fascia. This necessitates transposing it laterally and more superficial to make it more easily accessible for dialysis cannulation. Because of this, venous stenosis can occur at the angle of transposition where the vein transitions from deep to superficial since the more proximal portion of the vein is bound down by the brachial fascia (image 16). The stenosis is due to turbulence, although surgical manipulation and mechanical factors related to the brachial fascia may also contribute. This lesion that results has been referred to as the brachial-basilic angle of transposition (BAT) [104].
The frequency with which the BAT stenotic lesion is observed varies depending upon local practices since it is at least in part iatrogenic; however, it has been reported to be the most common complication associated with a brachial-basilic AV fistula, accounting for 87 percent of all interventions on this type of access [105,106]. Repeat intervention is frequently necessary. In one report, more than 50 percent of the cases required >3 angioplasties at the same site [105]. Elevated AV fistula intraluminal pressure distal to this stenosis results in luminal distention and aneurysm formation. These changes tend to be progressive and can ultimately lead to thrombosis and even loss of the access.
There have been very few studies directed specifically toward management of BAT stenosis. In one study of 85 transposed brachial-basilic fistulas, venous stenosis occurred in 54 percent. Of these, 74 percent qualified as BAT stenoses [105]. When treated with angioplasty, the one-year primary patency was 42 percent. Cumulative patency rate was 68, 58, and 53 percent at one, two, and three years, respectively. Repeat intervention was frequently necessary. Over half of the cases required more than 3 angioplasties with 29 percent requiring four or more interventions. The median time between angioplasties was 75 days.
In a study designed to evaluate the effect of stent placement for BAT restenosis, 37 cases were studied [104]. The cases served as their own controls with analysis of lesion and access patency both before and after stent placement. The pre-stent lesion primary patency rates at 3, 6 and 12 months were 55, 29, and 3 percent, respectively. Post-stent primary patency rates for the same periods were significantly higher at 69, 57, and 40 percent. The assisted primary patency rates for the same periods were 64, 39, and 13 percent (pre-stent) compared with 94, 91, and 80 percent (post-stent), which were also significantly higher.
Cephalic arch stenosis — The proximal cephalic vein is characterized by a swing point configuration that occurs as the cephalic arch passes though the coracoclavicular ligament just before it joins the subclavian vein. The development of cephalic arch stenosis has been attributed to hemodynamic factors related to vessel configuration, increased blood flow rate, external compression by accompanying structures, and hypertrophy of valves that are typically present at this location [5,107,108].
Both the brachial-cephalic and radial-cephalic AV fistulas deliver blood to the central circulation through the cephalic arch; however, cephalic arch stenosis is much more common in association with brachial-cephalic AV fistulas (39 versus 2 percent in one study [109], and 77 versus 20 percent in another [110]). Cephalic arch stenosis (image 17) is the most common stenosis seen in association with a brachial-cephalic AV fistula. In one study conducted over a five-year period, 77 percent of patients with a brachial-cephalic AV fistula who had a clinical indication for a venous angiogram (dysfunction) had cephalic arch stenosis [111].
Although angioplasty is the accepted initial treatment of venous stenosis associated with AV access, angioplasty of the cephalic arch is problematic because lesions at this site are more resistant to dilation, complications are more likely, and patency is reduced compared with other sites [107]. In a study of 177 cases of dysfunctional AV fistulas, 15 percent were due to cephalic arch stenosis [109]. Among 50 angioplasty procedures, anatomic and clinical success rates were 76 and 98 percent, respectively. Higher inflation pressures (>15 atm) were required in 58 percent of the cases. Rupture occurred in 3 of 50 patients, one of which led to AV fistula loss. Primary patency was 42 percent at six months and 23 percent at one year. Primary assisted patency was 83 percent at six months and 75 percent at one year. An average of 1.6 procedures per year was required per access.
Angioplasty using a cutting balloon has been studied as a method that might offer better results when dealing with resistant stenotic lesions. In a study of 74 dysfunctional brachial-cephalic AV fistulas, 30 (41 percent) were caused by cephalic arch stenosis [25]. Twenty-five procedures in 17 fistulas with this lesion were treated using a cutting balloon. In 15, a cutting balloon was used primarily, and in 10 procedures, use of the cutting balloon was followed by a standard balloon or high-pressure balloon angioplasty. Primary patency rates at 3, 6, 12, and 15 months were 94, 81, 38, and 22 percent, respectively. Assisted primary patency rates at the same intervals were 100, 94, 77, and 63 percent, respectively. When these results are compared with the one-year results reported in the study cited above [109], primary patency with the cutting balloon was better (38 versus 23 percent), but assisted primary patency was not (77 versus 75 percent).
Due to recurrent problems with angioplasty of the cephalic arch, the placement of a stent has been used to salvage the access. No randomized trials are available comparing angioplasty with stent placement for this lesion. One trial compared stent devices, randomly assigning 25 patients with recurrent cephalic arch stenosis associated with a brachial-cephalic fistula to either a bare-metal stent or a covered stent [112]. Among the 21 patients available to undergo angiography at three-month follow-up, the incidence of stenosis ≥50 percent was significantly higher for bare-metal stents compared with covered stents (70 versus 18 percent). Unfortunately, one half of the patients were lost to longer-term follow-up.
Surgical alternatives for treating cephalic arch stenosis include blood flow reduction (banding), stenotic segmental resection, cephalic-jugular vein bypass graft, cephalic vein transposition with venovenostomy, and basilic vein transposition with venovenostomy [113,114]. Because the abnormalities at the cephalic arch are thought to be related to increased flow, it has been postulated that flow reduction may be beneficial [115]. In a retrospective study of 33 patients who had undergone two or more treatments for cephalic arch stenosis, banding of the fistula reduced the need for further intervention [113]. Angioplasty rates for cephalic arch stenosis were calculated for the pre- and post-banding periods. The cephalic arch intervention rate was reduced from 3.34 to 0.9 interventions per access-year over an average follow-up of 14.5 months.
Taking advantage of the length of the cephalic vein in relationship to the basilic vein, cephalic vein transposition with the creation of a venovenostomy to the proximal basilic vein has proven to be very effective in the treatment of cephalic arch stenosis. In a study in which seven patients with brachiocephalic fistula and cephalic arch stenosis underwent a transposition procedure, primary patency was 70 percent at six months and 60 percent at 12 months [116]. The venovenostomy anastomotic site can also develop stenosis requiring treatment, but it responds much better than the cephalic arch lesion. One study evaluated outcomes of angioplasty of the cephalic arch lesion before and after transposition surgery in 13 patients [117]. Following transposition, all patients needed angioplasty. Angioplasty primary patency rates before surgical revision were 23, 8, and 0 percent at 3, 6, and 12 months, respectively. Cumulative patency rates for the same intervals were 100, 39, and 8 percent, respectively. After surgical transposition, the cumulative patency rate was 92 percent at 3, 6, and 12 months. Patients required significantly fewer interventions (1 intervention per patient-year following surgical transposition compared with 3.5 prior to).
Lesions unique to AV grafts — Two lesions that are unique to the AV graft include venous anastomotic stenosis, which is often resistant to conventional angioplasty techniques, and intragraft stenosis, which can usually be treated effectively without using alternative techniques.
Venous anastomotic stenosis — The most common site for stenosis associated with an AV graft is at the venous anastomosis (image 18). Approximately two thirds of reported venous stenosis lesions occur at the anastomosis [15]. The interface between the graft and the vein is a site of turbulent flow, which is thought to be due to the mismatch that occurs between the graft and the vein in terms of diameter and material compliance. This lesion may also be associated with stenosis extending into the first portion of the outflow vein.
Venous anastomotic lesions are relatively resistant to balloon angioplasty, often requiring the use of ultrahigh pressure angioplasty balloons for successful treatment. Technical success rates tend to be lower for angioplasty at this site compared with most other sites, and the recurrence rate is high [15,78,118-121]. In the control arm of one trial, the lesion primary patency for angioplasty at the venous anastomotic site was 20 percent at six months while that for the circuit was 38 percent [122]
Stenosis of the venous anastomosis is frequently treated following a thrombectomy procedure. In this subgroup, patency following angioplasty is even lower. In a study of 34 patients, the primary patency at the treated sites was 3 percent at six months, and assisted primary patency was 38 percent [123].
There are several alternatives to conventional angioplasty for venous anastomotic stenosis. Abandoning the AV graft in favor of a secondary AV fistula is the alternative preferred by the author in these cases. (See "Hemodialysis access following a failed arteriovenous access".)
The most intensively studied alternative to angioplasty for the treatment of AV graft venous anastomotic stenosis has been primary placement of covered stents (ie, in contrast to treatment for restenosis). A series of randomized trials have demonstrated that covered stents significantly improved patency compared with angioplasty alone.
●In a trial involving AV grafts, 190 patients with venous anastomotic stenosis were randomly assigned to conventional balloon angioplasty alone or balloon angioplasty plus placement of a covered stent [122]. At six months, lesion primary patency was significantly higher for covered stents compared with the balloon angioplasty (51 versus 23 percent), as was access primary patency (38 versus 20 percent). However, there was no significant difference in the rate of thrombosis between the two groups (33 versus 21 percent).
●A second study used the same covered stent but with a longer follow-up period of 24 months [124]. A total of 191 patients with AV graft anastomotic stenosis were randomly assigned to conventional balloon angioplasty alone or balloon angioplasty plus placement of a covered stent. Lesion primary patency rates for angioplasty plus covered stent were higher compared with angioplasty alone at 12 and 24 months (47.6 versus 24.8, and 26.9 versus 13.5 percent, respectively). Circuit primary patencies for the same time periods were 24 versus 11 percent, and 9.5 versus 5.5 percent, respectively. Restenosis requiring re-intervention occurred in 63.0 percent of the covered stent group during the 24 month follow-up and 82.6 percent of the angioplasty-alone group.
●In another trial, 293 patients with stenosis at the venous anastomosis of either dysfunctional (stenotic, n = 164) or thrombosed (n = 129) AV accesses were randomly assigned to balloon angioplasty alone or balloon angioplasty plus placement of a covered stent [125]. The six-month lesion primary patency was significantly improved for the covered stent group (51.6 versus 34.2 percent). The median time from the index procedure to the next intervention on the target lesion was significantly longer in the covered stent group (203 versus 108 days). Lesion primary patency for patients with a dysfunctional AV fistula was better compared with patients who had a thrombosed access for both treatments (covered stent 64.6 versus 36.1 percent; angioplasty alone, 45.8 versus 23.5 percent).
Other alternatives to conventional angioplasty that have also been evaluated for this lesion include cutting balloon angioplasty, use of drug-eluting balloons, and surgical treatment of the lesion. In a multicenter trial in which patients were randomly assigned to conventional angioplasty or cutting balloon angioplasty for a variety of lesions, outcomes were significantly improved for the cutting balloon group [126]. Similarly, the use of drug-eluting balloons has also been shown to benefit the treatment of venous anastomotic lesions. In one small trial of 40 patients (mixed cohorts of juxta-anastomotic AV fistula stenosis and AV graft venous anastomosis stenosis), primary patency at six months was significantly higher for the drug-eluting balloon group compared with angioplasty (70 versus 25 percent), as was cumulative patency (65 versus 20 percent) [46].
The results of patch angioplasty (surgical) may be inferior to covered stent placement. A retrospective review compared patch angioplasty with conventional balloon angioplasty plus covered stent placement in upper arm AV fistulas or graft [127]. Primary patencies for patch angioplasty compared with placement of a covered stent at one, three, and six months were 60 versus 68 percent, 33 versus 41 percent, and 13 versus 18 percent, respectively. Assisted primary patencies were 100 versus 80 percent, 67 versus 80 percent, and 47 versus 40 percent, respectively, at one, three, and six months.
Intragraft stenosis — The reported incidence of intragraft stenosis (image 19) varies widely, ranging from 2 to almost 40 percent [15,74,78]. The reasons for such variability are not clear but may be related to differences in the age of the AV grafts that were being studied. Unlike stenosis occurring at other sites, these lesions are not related to neointimal hyperplasia. They are caused by deposition of fibrin, lipid, and cellular debris within the graft. In addition, it is likely that defects in the graft from repeated needle puncture allow for some tissue ingrowth [128].
Intragraft stenosis is not uniformly distributed. In loop AV grafts, intragraft stenosis occurs primarily on the venous side of the loop, corresponding to cannulation areas. In straight grafts, the involvement generally spares the portion of the graft immediately adjacent to the arterial anastomosis [74].
Angioplasty has been widely used to treat these lesions. The mechanism by which the treatment results are obtained is different from lesions occurring at other sites. The expanded polytetrafluoroethylene (ePTFE) graft material cannot be dilated. Thus, the pressure exerted by the balloon squeezes, compresses, and perhaps dislodges the obstructing material. There are no reports in which results for this specific lesion following angioplasty were recorded except for an early study in which the 60, 180, and 360 day primary patency rates following angioplasty were 91, 76, and 55.9 percent, respectively [74]. If there are graft irregularities that are not responsive to angioplasty, or if the patient has frequent clotting events related to intragraft stenosis, the graft cannulation segment should be surgically replaced.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Dialysis" and "Society guideline links: Hemodialysis vascular access".)
SUMMARY AND RECOMMENDATIONS
●Indications for intervention – Endovascular intervention, primarily angioplasty, is the recommended initial treatment for stenotic lesions in both arteriovenous (AV) fistulas and AV grafts. To qualify for treatment, a stenotic lesion must be determined to be significant, which is defined as a narrowing of the vascular lumen equal to or greater than 50 percent, and associated with clinical symptoms, abnormal physical findings, and/or abnormal blood flow measurements. (See 'Endovascular intervention' above.)
●Angioplasty – Angioplasty uses an intravascular balloon positioned within a stenosis to dilate the vessel. Some type of anatomic disruption, such as a break or tear, accompanies a successful angioplasty. Anatomic success is defined as no more than 30 percent residual stenosis remaining following angioplasty (as compared with the adjacent normal vein on imaging). In addition, the abnormalities evident in the detection of significant pathology that led to treatment (eg, physical examination, hemodynamic/physiologic) should no longer be apparent following treatment. Because any degree of residual stenosis has been associated with shortened primary patency, our anatomic criterion following angioplasty is no residual stenosis on imaging. (See 'Angioplasty mechanism' above and 'Judging the success of angioplasty' above.)
●Recurrent stenosis – Repeat intervention is frequently necessary. The expected six-month patency rate following angioplasty for AV access dysfunction should be at least 50 percent. Higher rates of restenosis have been reported in association with specific types of lesions, including cephalic arch lesions, juxta-anastomotic stenosis, and angle of transposition lesions. (See 'Restenosis and repeat angioplasty' above and 'Sites of stenotic lesions' above.)
●Role of stents – Although stents can help manage failed angioplasty and other complications of endovascular intervention, they are sometimes inappropriately used. In addition, the use of stents adds cost and increases the risk for complications. The generally accepted indications for stent placement for treating hemodialysis AV access stenosis include acute angioplasty failure, rapid recurrence, and vein rupture following angioplasty. In general, patency rates following primary venous stenting are relatively low; however, for some lesion sites, the threshold for stent placement is lower, given a high risk for complications with repeated angioplasty and lack of surgical access. (See 'Stenting' above and 'Central vein stenosis' above.)
●Referral for surgical repair – For patients who are good surgical candidates, we agree with recommendations that state that if angioplasty of the same lesion is required more than two times within a three-month period, the patient should be considered for surgical revision as an alternative to placement of a stent. (See 'Surgical revision' above and 'Sites of stenotic lesions' above.)
●Outcomes – The outcomes of endovascular intervention for specific lesions are presented above. (See 'Specific AV access lesions' above.)
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟
