Abstract
Vascular access is the sine qua non of hemodialysis; it is also the Achilles heel. In this chapter, we review the different types of vascular access, including merits, disadvantages, and the lifecycle of an arteriovenous access, and describe current treatment for vascular access dysfunction.
Keywords
arteriovenous fistula, arteriovenous graft, central venous catheter, dialysis vascular access, fistula maturation, HeRO graft
Outline
Arteriovenous Fistula, 361
Types of Arteriovenous Fistulas, 361
Classification of Fistulas, 361
Fistula Types Based on Anatomy, 362
Life Cycle of the Arteriovenous Fistula, 362
Complications Associated with Arteriovenous Fistulas, 364
Failure to Mature, 364
Late Arteriovenous Fistula Failure, 365
Excessive Flow, 365
Secondary Arteriovenous Fistulas, 368
Arteriovenous Graft, 369
Complications of Arteriovenous Grafts, 370
Venous Stenosis, 370
Hemodialysis Reliable Outflow Vascular Access Device, 372
Dialysis Catheters, 373
Acute Dialysis Catheters, 377
Vascular access is the sine qua non of hemodialysis; it is also the Achilles heel. The process of cleansing the blood of the toxic products of chronic renal failure requires, as its first step, access to the circulation. The importance of the vascular access in the long-term provision of hemodialysis has been apparent from the beginning of the dialysis era. In 1944 Willem Kolff reported, after his first patient had received 12 dialysis treatments over a 26-day period, that all of the veins were “ruined.” A surgical cutdown of the radial artery followed but caused severe bleeding during heparinization. Kolff concluded that “we believe to be able to keep patients suffering from uremia and anuria alive so long as blood vessels for puncture are available.” It is interesting that more than six decades later we are still in the same position.
Fortunately, today we have better alternatives to provide access to the circulation; yet, it is unfortunate that none of these are ideally suited for the purpose. The ideal vascular access should have the basic characteristics listed in Box 23.1 .
Easy repetitive access to the circulation
Ability to reliably provide blood flow > 500 mL/min
Hemostasis at the end of dialysis accomplished easily and quickly
Function for the life of the patient
Freedom from complication
Cosmetically acceptable
Unfortunately, the ideal vascular access does not exist. There are several choices, which are classified as arteriovenous fistula (AVF), arteriovenous graft (AVG), or central vein catheter. Although each of these has its own set of problems, the characteristics of the AVF make it the preferred choice for most patients.
Arteriovenous Fistula
There are a number of benefits associated with an AVF. This type of access is associated with the best primary patency rate, the best cumulative patency rate, and the lowest rate of thrombosis and requires the fewest interventions over the duration of its life cycle. Cost of implantation and maintenance of an AVF is the lowest of the three types of access, and AVFs are associated with lower hospitalization rates than are seen in patients with AVGs (relative risk [RR] = 1.47) or catheters (RR = 2.3). Because patients with an AVF have fewer access-related problems, especially infection, they have a lower mortality and morbidity compared with patients with either an AVG or a central venous catheter.
With the exception of transposed fistulas, an AVF can be created with very little patient morbidity associated with the procedure compared with the insertion of a synthetic AVG, and the surgical procedure is relatively simple and is generally quickly accomplished.
Types of Arteriovenous Fistulas
Classification of Fistulas
AVF nomenclature incorporates its anatomical location, inflow artery, and outflow vein, such as radial-cephalic (the radial artery and the cephalic vein). Although there are standard AVF configurations, an AVF can be created anywhere there is a cannulatable vein (which may require transposition) of sufficient diameter to be attached to an adequate inflow artery. AVF types can be classified into three different categories: simple direct, vein transposition, and vein translocation ( Table 23.1 ).
Simple direct | Easiest of the AVFs to create. |
Vein and the artery are used in their normal positions. Distal end of the vein is freed and connected to the adjacent artery. | |
Vein transposition | Most difficult to create. |
Created using vein that is not easily assessable for dialysis use. Proximal end of the vein is left intact. Distal portion of the vein is transposed to an assessable position. Requires creation of a tunnel or pocket for the newly positioned vein. | |
Vein translocation | Vessel is removed from one anatomical location and moved to a new site. |
Requires the creation of both a venous and an arterial anastomosis. Requires the creation of a tunnel for its new location. |
Bioengineered veins have been developed and are currently in phase III clinical trials. The use of these vessels presents a model very similar to a transposition AVF.
Fistula Types Based on Anatomy
Although a variety of different anatomical types of AVF can be surgically created, most AVF creations fall within three basic configurations ( Fig. 23.1 ). There is a separate category of AVFs sometimes referred to as middle-arm or bidirectional fistulas that have been created when neither of the three basic configurations are possible. These are created using the proximal radial artery with either the median antebrachial vein or the median cubital vein. The suitability of the median antebrachial vein may be contributed to in part by the fact that the venipuncturist does not commonly access it.
According to accepted guidelines, the order of preference for the creation of permanent vascular access should start distally in the upper extremity and progress proximally to preserve venous anatomy. According to this approach, the radial-cephalic AVF would be considered primary and the brachial-cephalic and brachial-basilic would be considered secondary choices. Some have advocated that the middle-arm AVFs should be considered tertiary. If it is not possible to create one of the basic configurations of AVF, then reasonable attempts at creating a transposed AVF should be made before consideration is given to the insertion of AVG; however, it is important to consider patient-specific factors (e.g., comorbid conditions, anticipated longevity) at the time a surgical plan is devised. In the event that an AVG is placed, it should ideally be done with the plan that it will be used for the dual purposes of providing dialysis access during its problem-free life and for the development of the upper arm veins for the later creation of a secondary AVF once its use becomes problematic.
Life Cycle of the Arteriovenous Fistula
The life cycle of an AVF can be divided into five distinct clinical phases ( Fig. 23.2 ). The first three phases can be characterized as developmental stages ( Fig. 23.3 ) of the AVF, eventually leading to a clinically functional dialysis access. Although in most successful cases, an AVF progresses through these developmental stages over a period of 4 to 6 weeks, time is not considered an element of these definitions because it does not exert an influence on whether the defined outcome has been achieved.
Phase 1: Creation
Phase 1 of the AVF’s life cycle corresponds to the first developmental stage of the access—that is, creating an arteriovenous communication. If this is successful, as demonstrated by the presence of blood flow after creation, a patent AVF has been achieved.
Phase 2: Maturation
The maturation phase corresponds to the second developmental stage of the AVF and is characterized by evolution from AVF patency to a physiologically mature AVF, considered to have the potential for being used as a hemodialysis access. As stated earlier, AVF maturation is characterized by a continuous, progressive, and relatively rapid increase in blood flow and vessel diameter sufficient to permit reproducible clinical usability for hemodialysis. Studies have found that blood flow alone, or a combination of blood flow and vessel diameter, are reliable indirect indicators for predicting the successful use of an AVF for hemodialysis. Using these metrics, the physiologically mature AVF is defined as having an internal diameter >0.5 cm (measured without a tourniquet) and a blood flow >500 mL/min. A blood flow of 400 to 500 mL/min (measured from the brachial artery) has been found to have an accuracy of 53% to 93%, sensitivity of 67% to 96%, and specificity of 65% to 90% for predicting clinical AVF maturation. The combination of both AVF blood flow and internal vessel diameter of the vessel (500 mL/min, 0.5 cm) has been found to have a sensitivity of 84% and a specificity of 93%. It has been reported that most AVFs that mature will do so in 4 to 8 weeks.
Phase 3: Clinical Use, Initial
This third phase requires that successful clinical use of the AVF be demonstrated. This is the third and final stage of AVF development, referred to as a clinically functional AVF . This definition is necessary to separate from the second stage, AVF maturation, because clinical functionality is dependent on individual patient characteristics in addition to physiological changes that occur within the target vessel. It is possible for an AVF to be physiologically mature and have the capacity to provide adequate dialysis access but be unusable because of its location or depth. In addition, predialysis placement of the access will predictably result in a fully mature (physiologically) AVF not being used for a prolonged period and in some cases never being used. In these cases, proof of functionality is never obtained, or at least is delayed, even though optimal physiological changes may have occurred.
A clinically functional AVF is defined as one that can be cannulated with two dialysis needles for at least 75% of dialysis sessions within a 4-week period and achieves the prescribed dialysis. This definition, based on less than 100% success at cannulation, takes into account the fact that many patients with newly created AVFs have cannulation-related complications, although the access is destined to have long-term successful use.
Phase 4: Clinical Use, Sustained
Once the clinical functionality of an AVF has been proven, it enters the fourth phase of its life cycle, characterized by continuous, effective, problem-free use for hemodialysis. This is the ultimate criterion for judging success. Unfortunately, most cases will alternate between this phase and phase 5, characterized as dysfunction. The duration of this phase is indeterminate and is limited by the occurrence of problems and complications.
Phase 5: Dysfunction
This phase is characterized by the occurrence of a problem that interferes with the routine use of the AV access, threatens patency or results in a loss of patency, presents a significant risk for medical complication, or adversely affects the patient’s sense of well-being. Intervention is often required to resolve the problem. If this is successful, the AVF returns to phase 4.
Complications Associated with Arteriovenous Fistulas
Although the AVF is associated with fewer complications than are seen with other types of vascular access, they do occur and they should be dealt with effectively. The major complications that occur in conjunction with arteriovenous AVFs can be categorized under the headings of early failure, late failure, excessive flow, ischemia, aneurysm formation, and infection. Both early and late failures have multiple causes.
Failure to Mature
As stated earlier, an AVF is superior to an AVG; however, data on which AVFs’ superiority has been established are based on AVFs that successfully mature and become clinically functional and generally exclude AVFs that failed to mature. Were data from all AVFs (nonmaturing and maturing) to be considered, more than a year would elapse before the superiority of the AVF became apparent. This difference is due primarily to early failure. Failure to mature, also referred to as early or primary failure , is defined as an AVF that is never usable for dialysis or that fails within 3 months of use, and failure rates ranging from 20% to 60% have been reported. According to US Renal Data System data, 36% of AVFs created in 2014 failed and the average time between AVF creation and first use was 133 days. Although many AVFs can be salvaged, those that are often require more than one procedure to become clinically usable and have a shortened primary patency rate, making repetitive interventions necessary for continued clinical use.
Reasons for the high incidence of early AVF failure are not totally clear. However, one must realize that the dialysis practice and population at risk for end-stage renal disease (ESRD) have changed over time. When the radial-cephalic AVF first was described in 1966 by Brescia et al., nearly all patients had chronic glomerulonephritis, the average patient age was 43 years, and blood flows used for dialysis were 250 to 300 mL/min. The early AVF failure rate was 11%. Today’s dialysis patients are different—they are much older (many >70 years), and three-quarters of them have five or more comorbidities, with 90% having cardiovascular disease and 50% having diabetes. In addition, the average blood pump speed is much higher: 350 to 450 mL/min. Therefore it is not surprising that achieving functional AVFs in today’s ESRD population is often a challenge.
Additional risk factors for poorer AVF development include female sex, African American race, older age, greater body mass index (>35 kg/m 2 ), and history of diabetes, peripheral vascular disease, or coronary artery disease.
The three most commonly created AVFs, the radial-cephalic, the brachial-cephalic, and the brachial-basilic transposition, have differing rates of early failure. The early failure rate for the brachial-basilic transposed AVF is reported to be the lowest, followed by the brachial-cephalic and then the radial cephalic, which has the highest failure rate. The early failure rate for brachial-basilic transposition AVFs it is reported to be between 0% to 21% in various series. Compared with brachial-cephalic AVFs in the same series, brachial-basilic transpositions have an early failure rate of 0% versus 27%, 21% versus 32%, and 18% versus 38%. Radial-cephalic AVF failure rates of more than 60% have been reported.
Most investigators agree that nonmaturing AVFs have an associated anatomical problem. With the exception of thrombosed AVFs, three important principles have been established related to early failure: (1) A distinct lesion or lesions can generally be identified as the underlying cause of failure; (2) the problem can generally be identified by physical examination and confirmed by imaging; and (3) the lesions can be corrected with a high expectation of success (except for certain preexisting lesions, which should have been avoided by quality vascular mapping).
Although several problems commonly exist when an AVF fails to mature, the most common lesion observed in these cases is juxta-anastomotic stenosis. This is defined as stenosis occurring within the first 3 to 4 cm of the AVF, immediately adjacent to the arterial anastomosis ( Fig. 23.4A ). The anastomosis may also be involved, resulting in luminal narrowing, decreased AVF blood flow leading to problems of maturation, and early thrombosis.
In some instances, the problem is poor preoperative vessel selection. The cause of failure to mature may be related to lesions that should have been recognized before the creation of the access. In reported cases, preexisting proximal venous stenosis has been documented as present in 4% to 59% of cases and central venous stenosis in 2.6% to 9% of cases. Small arteries or arteries with stenotic lesions may be present and contribute to AVF nonmaturation with an incidence ranging between 4% to 6%. When an AVF is created, the inflow artery generally dilates in parallel with dilatation and increasing blood flow in the AVF. If this fails to occur, the AVF will likely not mature.
Another type of preexisting problem that can affect AVF development and maturation is the presence of accessory veins. These are side branches of the forearm veins used for the construction of an AVF ( Fig. 23.4B ). Most accessory veins are not problematic; however, because AVF maturation is dependent on blood flow, a large accessory vein that diverts a major portion of the flow away from the target vessel can result in failure of maturation.
Late Arteriovenous Fistula Failure
Late AVF failure is defined as failure that occurs after a period of normal use. The primary causes of late failure are venous stenosis and acquired arterial lesions ( Fig. 23.5 ). These lesions are manifest as a pathophysiological change in the AVF resulting from increasing resistance, leading to a decline in blood flow, followed by inadequate dialysis and eventually thrombosis. The same types of lesions that are seen in association with early failure may be also seen here. Whether present initially and clinically unimportant or whether they developed (or progressed) over time during AVF use is not clear.
The most common cause of late AVF failure is venous stenosis. The site of the relevant lesion varies with the site of the AVF arteriovenous anastomosis. In distal radial-cephalic AVFs, virtually all stenoses are found in the inflow region (anastomotic and juxta-anastomotic), whereas outflow lesions are found almost exclusively in midforearm and elbow/upper arm AVF. Venous stenosis associated with an AVF generally develops at areas of vein bifurcation, at swing points, and in association with venous valves. The development of collateral veins is common, often extensive, and tends to preserve flow in the access ( Fig. 23.6 ).
Stenosis of the inflow artery has been reported as a cause of late AVF failure in 6% to 18% of cases. These lesions can also lead to decreased blood flow in the access, resulting in inadequate dialysis and eventually thrombosis. In general, 100% of all thrombosed AVFs have either venous or arterial-associated pathological anatomy. Thrombosis is the endpoint of late AVF failure, and it occurs at a rate that is approximately one-sixth of that for an AVG.
Excessive Flow
After the early, rapid increase in dialysis access blood flow (Qa) after AVF creation, there is a tendency for Qa to continue to slowly, yet progressively increase for the life of the access. Although this is generally tolerated and in some cases enhances the dialysis functionality of the AVF, it can also lead to problems, the most clinically relevant of which is high-output heart failure (HOHF). HOHF is a particularly serious problem considering the fact that cardiovascular disease is the leading cause of death in the ESRD population. Currently, there is no generally accepted definition of excess Qa. It has been suggested that the ratio of Qa to cardiac output (Qa/CO), referred to as cardiopulmonary recirculation (CPR), be used as a gauge. Normal values for this ratio are generally considered to be in the range of 20% to 25%. Patients with levels greater than this, or with Qa in excess of 2 L/min ( Fig. 23.7 ), should be considered at higher risk for the development of HOHF.
The risk for HOHF is directly proportional to the Qa and inversely proportional to the patient’s preexisting cardiac status. Many patients can tolerate a high Qa associated with an elevated CPR; however, even levels considered to be within the normal range can be excessive for the impaired heart because compromised cardiac function may limit the increase in CO that is required. For this reason, it has been suggested that the cardiac index may be a better gauge for the establishment of HOHF because it is more individualized. Moreover, it has been suggested that patients with symptoms of heart failure (dyspnea either at rest or with varying degrees of exertion; orthopnea; paroxysmal dyspnea; and edema, pulmonary and/or peripheral) that are resistant to medical management and occur in association with a cardiac index >4.0 L/min/m 2 should be considered as having HOHF.
Hand Ischemia: Dialysis Access Steal Syndrome
When an arteriovenous access is placed in an extremity, a unique physiological state is established, because blood flow to the hand must now also supply the access. This hand/vascular access complex consists of a proximal artery feeding into two competing circuits connected in series—the arteriovenous access and the peripheral vascular bed on which perfusion of the hand is dependent. The former is a low-resistance pathway and is located proximally. The latter is downstream and is characterized by high resistance. Collateral arteries that bypass the access circuit to feed the periphery directly also contribute to this complex.
Most patients tolerate this abnormal physiological state because of compensatory mechanisms. However, in some instances, there is a failure of these mechanisms and hand ischemia, referred to as dialysis access steal syndrome (DASS), occurs. DASS can occur with either an AVG or an AVF; however, it is more common with the latter. Two distinct clinical variants of hand ischemia are recognized as associated with the placement of a dialysis access: ischemic monomelic neuropathy (IMN), where changes are confined to the nerves of the hand, and DASS, in which ischemic changes affect all tissues of the hand to a varying degree of severity.
DASS is reported to occur in 1.6% to 8% of cases with an arteriovenous access. Major predisposing risk factors include the use of the brachial artery as the inflow, diabetes, female sex, age >60 years, peripheral artery disease, and multiple previous access procedures. Of these, the use of the brachial artery appears to create the greatest risk for this condition. In addition, DASS is reported to occur more readily in patients with AVFs with large anastomoses and Qa.
Based on patient-specific clinical signs and symptoms, DASS is categorized into four stages reflecting increasing clinical severity ( Box 23.2 , Fig. 23.8 ). These four stages guide treatment choices for DASS. DASS varies in its time of onset and may be acute, subacute, or chronic. Acute DASS is defined as the onset of signs and symptoms that appear immediately or within hours of the time of the surgical procedure. Acute DASS is most commonly associated with the placement of an AVG. DASS that occurs later, but within 1 month of access placement, is defined as subacute. Those cases occurring after this period are classified as chronic. These latter two groups are more commonly seen in association with AVFs and more specifically with brachial artery–based AVFs, although either an AVF or an AVG can be involved at any the three periods.
Stage I: Pale/livid hand and/or cool hand without pain
Stage II: Pain during exercise and/or during dialysis
Stage III: Rest pain or loss of motor function
Stage IV: Tissue loss
It is important to differentiate DASS from IMN. IMN is a distinct clinical entity associated with interference of a major limb artery resulting in multiple distal, axonal-loss mononeuropathies. The attributable causal factor is believed to be occlusion of the brachial artery during the AV access surgical procedure. The syndrome develops quickly, typically within minutes to hours of AVF creation. The pathognomic feature is the presence of diffuse neurological dysfunction, usually in the absence of significant ischemic changes in the tissues of the hand and fingers, which differentiates it from DASS. Attributable neuropathic symptoms include pain, paresthesias, and numbness in the distribution of all three forearm nerves along with diffuse motor weakness or paralysis. Involvement of fewer than all three forearm nerves should bring the diagnosis into serious question. Typically the hand is warm, capillary refill is preserved, and a palpable radial or ulnar pulse or audible Doppler signal is present. The motor deficits resulting from the IMN nerve damage cause severe disability to the involved hand ( Fig. 23.9 ).
The exact incidence of IMN is not known, because most of the literature is based on case reports, but it is not believed to be a common occurrence. IMN occurs exclusively in association with an arteriovenous access that is brachial artery based. There are no reports of IMN precipitated by a distal forearm procedure. Although there are case reports to the contrary, the condition occurs almost exclusively in diabetic hemodialysis patients.
Aneurysm Formation
Most of the bulges that are noted in association with an AVF are true aneurysms; that is, they contain all layers of the vessel wall. Although there is no universally accepted definition for an aneurysmal AVF, it is commonly defined as an AVF that is three times the diameter of the adjacent normal vein or a minimum of 2 cm. These pathological lesions represent degenerative changes in the vein wall and are relatively common. The features that characterize an aneurysm typically result from the combination of repeated dialysis needle punctures and hemodynamic factors, such as downstream peripheral or central venous stenosis. As an aneurysm develops and expands, it can lead to complications such as pain, objectionable cosmetic appearance, difficult cannulation, risk for bleeding, and problems with access blood flow.
Aneurysms can be fusiform, ectatic, or spherical ( Fig. 23.10 ). The fusiform type of aneurysm is commonly seen in association with segmental overuse—that is, repetitive cannulation in a localized area leading to weakening of the vessel wall. With increased pressure, progressive ballooning of the vessel wall develops, generally at the arterial and venous cannulation sites, and have a fusiform appearance. The ectatic type of aneurysm represents a relatively diffuse enlargement of the entire AVF related to downstream stenosis. If allowed to progress, it eventually achieves an appearance that has been referred to as a megafistula. The spherical aneurysm is related to a localized defect in the wall of the AVF. At times, this anomaly appears acutely in association with a problematic cannulation and may be a pseudoaneurysm, rather than a true aneurysm.
Infection
Infections in AVF are uncommon and are reported to occur in less than 0.4% of cases in the postoperative period and 0.2% per year thereafter. This is considerably less than that of AVGs and catheters. AVF-associated infection can take the form of cellulitis, an abscess, bacteremia, septic emboli, and sepsis. AVF infections occurring during the immediate postoperative period are generally related to a problem with aseptic technique during surgery, whereas those occurring later are most often related to contamination during cannulation. Late AVF infection has been attributed to the buttonhole cannulation technique.
Secondary Arteriovenous Fistulas
Secondary AVFs (SAVF) are important to the hemodialysis vascular access strategy of all nephrologists. A SAVF is defined as an AVF that is created after an AV access, usually an AVG, using the outflow veins of that access. In general, this is a forearm access and the AVF is created using an upper arm draining vein. Because of the forearm access, the veins of the upper arm undergo the same process of maturation as is seen with AVF development and for the same reasons. In most cases this involves the creation of either a brachial-cephalic or a brachial-basilic AVF. Even when one of these veins is not suitable, there may be an adequate vein in the forearm that can be used as a bidirectional, middle-arm SAVF.
When an AV graft is placed in a patient, it should be done with a dual purpose in mind—first, to provide an access for hemodialysis, and second, as a means of maturing veins in the upper arm for a SAVF. The National Vascular Access Improvement Initiative (Fistula First) recommended that every patient receiving dialysis via an AVG be viewed as a potential candidate for a SAVF. To identify a suitable patient for an SAVF, visualizing the veins of the upper arm is necessary and is most easily accomplished by simply having patients roll up their sleeve ( Fig. 23.11 ). With any type of procedure to treat AVG dysfunction, the angiographic studies that are performed can be used to identify optimal candidates for a SAVF if the operator is alert to the issue. In a study of the angiograms that were performed as part of either an angioplasty or a thrombectomy procedure, 75% of the patients with lower-arm grafts were found to have one or both upper-arm superficial veins that were optimal for SAVF conversion.
The success with SAVF creation has been very good, particularly considering that many of these patients were not initially suitable for AVF creation, and they often have a history of prior procedures. The 1-year primary and cumulative patency rates for SAVFs have been reported in the range of 71% to 82.5% and 92.5% to 100%, respectively.
Arteriovenous Graft
Although the AVF was introduced by Brescia and coworkers in 1966, an AVF was not possible in many patients. After that innovation, vascular access evolved through the use of saphenous vein translocation, bovine carotid artery graft, and the Dacron velour vascular graft, but still there were major short-term and chronic problems with these access types. In 1976 the problem of vascular access was solved, or so it was thought, through the use of expanded polytetrafluoroethylene (ePTFE). The AVG was initially considered only as a substitute access in patients in whom an AVF was not possible; however, this philosophy soon became lost and it became the access of choice in the United States, a practice that led to major problems and one that has been very difficult to reverse.
In 1997, the first iteration of National Kidney Foundation Kidney Disease Outcomes Quality Initiative (NKF KDOQI) Clinical Practice Guidelines for Vascular Access was published and stressed the need to decrease the reliance on AVGs. However, in 1998, 58% of patients on dialysis for 2 or more years were still using an AVG. The National Vascular Access Improvement Initiative, which came to be referred to as Fistula First, was launched in January 2003. This was a Medicare initiative designed to decrease AVG reliance and increase AVF use in the United States. Since that time there has been a slow but progressive decline in the prevalence of AVG use, declining to a level of 18.3% in 2014.
It is important to note that there remains a population of patients for whom an AVG represents the best choice for a dialysis vascular access. In general, these patients fall into two categories: (1) patients with vascular anatomy that is not favorable for the creation of an AVF as determined by vascular mapping; (2) patients with comorbidities that increase the risk for short life expectancy and the time required for an AVF maturation and for whom the high risk for primary failure makes AVF creation imprudent. Elderly patients also fall within this latter group. In a study using Medicare-derived data, dialysis survival was tabulated and four age groups were evaluated: 65 to 79, 80 to 84, 85 to 89, and >90. At 6 months after dialysis initiation, the mortality was 20%, 30%, 40%, and 45%, respectively; in accordance with these results, it was found that an AVF was not associated with better survival compared with an AVG used as the first predialysis access among patients older than 80 years.
Advantages of Arteriovenous Graft
Although considered more problematic than AVFs, AVGs have several advantages. They are comparatively easy to insert and repair, can be used at multiple anatomical sites, can be created in a variety of shapes and configurations, have a short maturation time, are technically easy to cannulate, and offer a large area for cannulation. Although there is a higher infection rate compared with an AVF, their primary disadvantage relates to the development of stenosis from neointimal hyperplasia, which tends to shorten their patency. Reported primary patency rates for AVGs range from 40% to 60% at 1 year, yet with aggressive management of thrombosis, secondary patency rates as high as 90% at 1 year can be achieved.
Types of Arteriovenous Grafts
Unfortunately, a large proportion of patients start dialysis with a central venous catheter, which creates an inordinate risk to the patient. Using the standard ePTFE graft shortens the time required to develop a clinically functional AVF; however, traditional dialysis programs wait 2 to 3 weeks before the use of a newly placed AVG, so a central venous catheter is still required for dialysis access. Data from the Dialysis Outcomes and Practice Patterns Study (DOPPS) was used to examine the average time of AVG cannulation. Investigators found that the relative risk for graft failure (reference group = first cannulation at 2 to 3 weeks) was 0.84 with first cannulation at <2 weeks ( P = 0.11), 0.94 with first cannulation at 3 to 4 weeks ( P = 0.48), and 0.93 with first cannulation at >4 weeks ( P = 0.48), suggesting that earlier cannulation was not associated with AVG failure. Moreover, multilayered early cannulation grafts have been developed that can be cannulated within 24 hours of placement and have no increase in complication rates compared with traditional ePTFE.
An AVG can be placed in a number of different sites and configurations, as long as the cannulation segment is accessible and there is blood return to the right atrium. However, the most common AVG sites are the forearm, the upper arm, and the thigh. Although it is distinctly unusual to see anything other than a loop graft in the thigh, either straight, curved, or loop grafts may be placed in the upper extremity locations.
Complications of Arteriovenous Grafts
Complications occur much more commonly with AVGs than with AVFs. The major complications are venous stenosis (including the venous anastomosis), arterial stenosis (including the arterial anastomosis) thrombosis, infection, pseudoaneurysm formation, and hand ischemia.
Venous Stenosis
The most common complications associated with an AVG are venous stenosis, arterial stenosis, and thrombosis. In most cases the problems of stenosis and thrombosis share the relationship of disease and symptom. The histological characteristics of venous and arterial stenosis are well characterized as aggressive neointimal hyperplasia. The net result of this process is a progressively expanding lesion that encroaches on the lumen of the vessel, causing a progressive increase in resistance and a decrease in flow, often resulting in eventual thrombosis.
Venous stenosis occurs most often at the venous anastomosis but may occur anywhere within the access conduit, composed of the arterial anastomosis, AVG, the venous anastomosis, and its peripheral and central draining veins. In a review of 2300 cases of venous stenosis, the following distribution of lesions was found: venous anastomosis 60%, peripheral vein 37%, within the graft 38%, central veins 3.2%, and multiple locations 31% ( Fig. 23.12 ).
Percutaneous angioplasty is the treatment of choice for venous stenosis. Among a pooled cohort of 2166 cases from 15 published studies, AVG-related peripheral vein stenosis treated with angioplasty was associated with a primary patency of 62% and a cumulative patency of 85% at 6 months. In metaanalysis based on 34 relevant studies, the primary and cumulative patency rate for AVGs at 6 and 18 months was 58% and 33% and 76% and 55%, respectively.
The dialysis vascular access should be thought of as a complete circuit starting and ending with the heart. The venous side represents only one-half of the circuit, whereas the other half is arterial. Arterial lesions adversely affect access inflow, and a hemodynamically significant lesion can develop anywhere in the arterial tree from the ascending aorta to the arterial anastomosis. Reports of inflow stenosis in dysfunctional AVGs range between 14% to 42%. These lesions are generally treatable with angioplasty.
Over time, progressive stenosis can result in thrombosis. The frequency of AVG thrombosis is approximately 1 to 1.5 per patient per year. Approximately 85% to 90% of all cases of graft thrombosis are associated with an underlying anatomical lesion.
Infection
Infection ( Fig. 23.13 ) is a common cause of AVG loss and a prominent cause of patient mortality. In one report, infection was reported to occur at a frequency of 1.3 episodes per 100 dialysis months and was associated with bacteremia at a rate of 0.7 cases per 100 dialysis months. In a prospective Canadian study in which surveillance for hemodialysis-related bloodstream infections was performed in 11 centers during a 6-month period, it was found that the relative risk for bloodstream infection with an AVG access was 2.5 per 1000 dialysis procedures. This was compared with a rate of 0.2 in patients receiving dialysis via an AVF.
The hospitalization rate for dialysis patients is double that of the general population, and infection is the attributable cause in 20% of cases. The development of dialysis access sepsis is associated with a higher incidence of myocardial infarction, congestive heart failure, cerebral stroke, and peripheral artery ischemic disease in the subsequent years. Not only does AVG infection result in significant morbidity, it results in AVG loss in a proportion of affected patients.
A number of risk factors for AVG infection have been recognized, including cannulation technique, AVG location, and duration of use. Patient personal hygiene appears to be the most important risk factor for the development of access-related infection ; however, episodes of infection have been traced to individual dialysis facility staff and can be related to poor needle insertion technique. This underscores the importance of staff training in infection control measures.
Femoral AVGs are often placed when all upper limb access sites have been used. These AVGs are at higher risk for infection. The incidence of graft infection increases with the duration of AVG use, suggesting another reason to evaluate AVG patients for a SAVF.
AVG infections are generally attributable to common skin microorganisms represented by gram-positive bacteria. In most cases, the causative organism of the AVG infection is Staphylococcus aureus or other gram-positive pathogens, such as coagulase-negative staphylococci. S. aureus is identified in nearly 68% cases of AVG infection, followed by Staphylococcus epidermitis. Gram-negative bacteria are less commonly the cause of infection, and some episodes are polymicrobial.
Pseudoaneurysm Formation
A pseudoaneurysm (“false aneurysm”) is characterized by actual disruption of the layers of the AVG leading to a bulging anatomical defect ( Fig. 23.14 ).