**Abstract**

Endovascular intervention in hemodialysis vascular access is among the most frequent interventions performed in an angiography suite. Vascular stenosis is the most prevalent lesion causing vascular access malfunction. Vascular access pathology and the outcomes in response to endovascular treatment are quite different from the arterial territory. Treatment strategy must be integrated, multidisciplinary, and with a long-term perspective, as recurrence rates of malfunction are quite common. We will detail our experience managing an extremely busy vascular access center serving a population of 4000 dialysis patients, performing all endovascular techniques in close coordination with the surgical team.

**Keywords:** hemodialysis, vascular access, angioplasty

#### **1. Introduction**

Endovascular interventions have substituted surgical repair as the primary treatment of failing or thrombosed vascular access (VA). Endovascular and surgical techniques, however, are complementary. Optimizing endovascular interventions of VA malfunction is a crucial component for a successful vascular access program. The identification and early treatment of stenosis are essential to prevent access thrombosis and ultimate failure.

Despite recent advances in endovascular techniques and devices, angioplasty continues to be the primary method for the treatment of access-related stenosis. Not all stenosis needs to be treated. When timely applied, angioplasty is a fast, easy, and safe procedure that can extend the patency of a hemodialysis graft or fistula.

The *early detection and endovascular intervention to correct dialysis vascular access malfunction* are reviewed in this chapter, describing the authors' experience in a highly active Vascular Access Center in Lisbon, integrated in a large outpatient dialysis network. We will cover the following topics: (1) vascular access options and its selection, (2) vascular access morbidity and complications, (3) vascular access malfunction detection, (4) endovascular interventions to correct dialysis vascular access malfunction, and (5) endovascular intervention outcomes.

#### **2. Vascular access options and its selection**

The VA constitutes the interface between chronic kidney disease (CKD) patient and machine (the dialysis monitor); its function is a key factor that affects most

dialysis treatment quality indicators, such as dialysis dose and adequacy (Kt/V), substitution volume during hemodiafiltration, operating costs, and vital prognosis of the dialysis patient.

In this chapter, we deal only with long-term arteriovenous accesses:


The VA dysfunction and its complications, such as low access flow (Qa), infection, loss of dialysis adequacy, or thrombosis, are the major single cause of hospitalization and morbidity requiring endovascular intervention, as well as one of the most important drivers of the total cost of an end-stage renal failure program.

Whenever a native arteriovenous fistula (AVF) can be built and is able to mature in no more than 8 weeks, it is considered the first and best choice as a vascular access. It results in higher long-term longevity and less thrombotic or infectious morbidity, needs fewer procedures for maintenance, and is overall a big life and money saver.

The nAVF, however, comes with its own set of disadvantages. There is a higher risk of primary failure (nonmaturation) up to 60% prior to cannulation, requiring frequent angiographic procedures to assist maturation [1–3]. Studies have shown that the primary failure rate is two times greater for fistulas (40%) than grafts (19%), with similar cumulative patency; in addition, the number of catheter days before AV access use was more than double in those using a fistula (81 days) than those with AV grafts (38 days); however, grafts require more angioplasties (1.4 vs. 3.2 events) and thrombolysis (0.05 vs. 0.98 events) interventions per 1000 patientdays [2, 4]. The risk of primary fistula failure is much higher for lower arm fistula (28%) than with upper arm fistula (20%), although these last ones produce more than 90% of all cephalic arch stenosis [1].

The secondary patency rates of AV grafts (total life span even if requiring several interventions to maintain its function) are on average around 3 years, all in all identical to AV fistulas, but those improved rates are achieved at the expense of three- to six-fold greater reintervention rates.

There has never been a randomized control trial (RCT) comparing different VA choices regarding mortality or other hard outcomes. All large observational trials compared accesses achieved as opposed to the accesses that were intended (as in intention to treat). As 25–60% of all AVFs created either fail or need several procedures to mature and the central venous catheter (CVC) group in most studies were people in whom AVF failed or CVC was chosen because of a predictable bad prognosis (age, congestive heart failure, short life expectancy, etc.), we really cannot answer the question on which VA is the best. If we exclude patients that begin hemodialysis urgently, mortality between nAVF and CVC patients becomes identical. Using a decision analysis model (fed with data extracted from DOPPS 2, the

**185**

societies in this field.

*Early Detection and Endovascular Intervention to Correct Dialysis Vascular Access Malfunction*

REDUCE FTM study, the DAC study, and CMS data) for choosing the best option for patients initiating hemodialysis (HD) with a CVC, a nAVF attempt strategy is associated with better survival and lower annual cost, but that advantage is progres-

In this chapter, we basically describe our experience on VA management in our dialysis network treating approximately 5000 patients in our Vascular Access Center (VAC) that performs more than 1000 VA surgeries and more than 1600

The most common VA complications are failure to mature, persistently low Qa, suboptimal dialysis adequacy, pain, aneurysms, rupture/hemorrhage, infection, and thrombosis. Endovascular stenosis is the underlying lesion and the direct

Neointimal hyperplasia is the common pathogenic mechanism inducing stenosis, and stenosis is the underlying promoter of thrombosis. Stenotic plaques are composed of myofibroblasts (smooth muscle cells) surrounded by extracellular matrix and macrophages. This cell proliferation begins in the adventitia and migrates toward the lumen of anastomotic areas or endothelial segments exposed to several stresses, such as surgical trauma, shear stress, wall stress, diameter and compliance mismatch, uremic endothelial dysfunction, and wall lesion secondary

Stenosis is necessary for thrombosis, but it is not enough. Only 30% of stenosis above 50% of lumen compromise will cause thrombosis in the next 6 months; we just do not know which ones. On the other hand, angioplasty induces accelerated NH with recurrent stenosis [7]. In 20% of the cases, recurrent stenosis occurs in 1 week and in 40% in 1 month [8], and although stenosis stenting may delay

As in other vascular territories, we do not know and have no biomarkers to decide which stenosis will progress to cause thrombosis, which stenosis if dilated will prevent thrombosis, which stenosis once dilated will suffer early recurrence, which is the best option to prevent recurrence, and how to define the successful angioplasty.

In hemodialysis vascular access management, just as in general medicine, an early diagnosis of malfunction and prevention of definite failure is considered the best approach to diminish morbidity and costs. This axiom was strongly suggested in several seminal studies [10, 11] and is expressed in most guidelines of scientific

It is recommended that regular monitoring of access function should be performed, preferably by measuring vascular access flow (Qa), and when access stenosis is present, preemptive intervention should be performed percutaneously without further delay. In support of these level 2 recommendations, we can quote: "All types of pressure measurement should be abandoned in favor of access flow

stenosis recurrence, it did not reduce the incidence of thrombosis [9].

Access malfunction is a source of tremendous emotional and physical suffering, dialysis treatments loss, low treatment adequacy, urgent need for a central catheter as a substitution access, and referral for new angiography or surgical procedures at

*DOI: http://dx.doi.org/10.5772/intechopen.92631*

endovascular interventions per year.

culprit behind most of these complications.

**4. Vascular access malfunction detection**

to repeated needle punctures.

huge costs.

sively lost in patients above 60 years or diabetics [5, 6].

**3. Vascular access morbidity and complications**

*Early Detection and Endovascular Intervention to Correct Dialysis Vascular Access Malfunction DOI: http://dx.doi.org/10.5772/intechopen.92631*

REDUCE FTM study, the DAC study, and CMS data) for choosing the best option for patients initiating hemodialysis (HD) with a CVC, a nAVF attempt strategy is associated with better survival and lower annual cost, but that advantage is progressively lost in patients above 60 years or diabetics [5, 6].

Access malfunction is a source of tremendous emotional and physical suffering, dialysis treatments loss, low treatment adequacy, urgent need for a central catheter as a substitution access, and referral for new angiography or surgical procedures at huge costs.

In this chapter, we basically describe our experience on VA management in our dialysis network treating approximately 5000 patients in our Vascular Access Center (VAC) that performs more than 1000 VA surgeries and more than 1600 endovascular interventions per year.

### **3. Vascular access morbidity and complications**

The most common VA complications are failure to mature, persistently low Qa, suboptimal dialysis adequacy, pain, aneurysms, rupture/hemorrhage, infection, and thrombosis. Endovascular stenosis is the underlying lesion and the direct culprit behind most of these complications.

Neointimal hyperplasia is the common pathogenic mechanism inducing stenosis, and stenosis is the underlying promoter of thrombosis. Stenotic plaques are composed of myofibroblasts (smooth muscle cells) surrounded by extracellular matrix and macrophages. This cell proliferation begins in the adventitia and migrates toward the lumen of anastomotic areas or endothelial segments exposed to several stresses, such as surgical trauma, shear stress, wall stress, diameter and compliance mismatch, uremic endothelial dysfunction, and wall lesion secondary to repeated needle punctures.

Stenosis is necessary for thrombosis, but it is not enough. Only 30% of stenosis above 50% of lumen compromise will cause thrombosis in the next 6 months; we just do not know which ones. On the other hand, angioplasty induces accelerated NH with recurrent stenosis [7]. In 20% of the cases, recurrent stenosis occurs in 1 week and in 40% in 1 month [8], and although stenosis stenting may delay stenosis recurrence, it did not reduce the incidence of thrombosis [9].

As in other vascular territories, we do not know and have no biomarkers to decide which stenosis will progress to cause thrombosis, which stenosis if dilated will prevent thrombosis, which stenosis once dilated will suffer early recurrence, which is the best option to prevent recurrence, and how to define the successful angioplasty.

#### **4. Vascular access malfunction detection**

In hemodialysis vascular access management, just as in general medicine, an early diagnosis of malfunction and prevention of definite failure is considered the best approach to diminish morbidity and costs. This axiom was strongly suggested in several seminal studies [10, 11] and is expressed in most guidelines of scientific societies in this field.

It is recommended that regular monitoring of access function should be performed, preferably by measuring vascular access flow (Qa), and when access stenosis is present, preemptive intervention should be performed percutaneously without further delay. In support of these level 2 recommendations, we can quote: "All types of pressure measurement should be abandoned in favor of access flow

*Cardiac Diseases - Novel Aspects of Cardiac Risk, Cardiorenal Pathology and Cardiac Interventions*

dialysis treatment quality indicators, such as dialysis dose and adequacy (Kt/V), substitution volume during hemodiafiltration, operating costs, and vital prognosis

**A.** The native AV fistula (nAVF), usually result from an end-to-side anastomosis of a vein to an artery, either at the wrist (distal fistula), most commonly a radio-cephalic fistula, or at the elbow/upper arm level (proximal fistula), in this position most commonly a brachiocephalic fistula, or a brachio-basilic fistula, this last one requiring a second procedure the transposition to the

**B.** The arteriovenous graft (AVG), usually a second choice in patients not suitable for a nAVF fistula, has better mechanical strength, can be used earlier, and has lower primary failure rates when compared with nAVF, but has a much higher infection risk, a poorer primary long-term patency, and needs many more

The VA dysfunction and its complications, such as low access flow (Qa), infection, loss of dialysis adequacy, or thrombosis, are the major single cause of hospitalization and morbidity requiring endovascular intervention, as well as one of the most important drivers of the total cost of an end-stage renal failure program.

Whenever a native arteriovenous fistula (AVF) can be built and is able to mature

The nAVF, however, comes with its own set of disadvantages. There is a higher risk of primary failure (nonmaturation) up to 60% prior to cannulation, requiring frequent angiographic procedures to assist maturation [1–3]. Studies have shown that the primary failure rate is two times greater for fistulas (40%) than grafts (19%), with similar cumulative patency; in addition, the number of catheter days before AV access use was more than double in those using a fistula (81 days) than those with AV grafts (38 days); however, grafts require more angioplasties (1.4 vs. 3.2 events) and thrombolysis (0.05 vs. 0.98 events) interventions per 1000 patientdays [2, 4]. The risk of primary fistula failure is much higher for lower arm fistula (28%) than with upper arm fistula (20%), although these last ones produce more

The secondary patency rates of AV grafts (total life span even if requiring several interventions to maintain its function) are on average around 3 years, all in all identical to AV fistulas, but those improved rates are achieved at the expense of

There has never been a randomized control trial (RCT) comparing different VA choices regarding mortality or other hard outcomes. All large observational trials compared accesses achieved as opposed to the accesses that were intended (as in intention to treat). As 25–60% of all AVFs created either fail or need several procedures to mature and the central venous catheter (CVC) group in most studies were people in whom AVF failed or CVC was chosen because of a predictable bad prognosis (age, congestive heart failure, short life expectancy, etc.), we really cannot answer the question on which VA is the best. If we exclude patients that begin hemodialysis urgently, mortality between nAVF and CVC patients becomes identical. Using a decision analysis model (fed with data extracted from DOPPS 2, the

in no more than 8 weeks, it is considered the first and best choice as a vascular access. It results in higher long-term longevity and less thrombotic or infectious morbidity, needs fewer procedures for maintenance, and is overall a big life and

In this chapter, we deal only with long-term arteriovenous accesses:

of the dialysis patient.

money saver.

surface of the arterialized vein.

interventions to remain functional.

than 90% of all cephalic arch stenosis [1].

three- to six-fold greater reintervention rates.

**184**

measurement," and "Monitoring plus intervention reduces thrombotic rates, morbidity and costs" [11, 12].

Consensual recommendations for preemptive intervention in malfunctioning grafts are (a) Qa measurement <600 ml/min for grafts or <400 ml/min for native fistulas and (b) a Qa drop higher than 25% over two consecutive measurements [13].

However, recent and quite relevant information has questioned those recommendations, and scanning through recent prospective randomized controlled trials in this field reveals some discordant opinions.

No matter if we are looking at native fistulas or PTFE grafts, using only Qa measurements, or its association with Doppler studies or dynamic venous pressure as surveillance techniques, it is believed that VA stenosis is now very effectively detected and responsible for a large increase in percutaneous vascular access procedures. Surprisingly, however, it has been found that all these diagnostic and therapeutic procedures fail to reduce the thrombosis rate or prolong access longevity, fueling an ongoing controversy regarding its beneficial effects, both in terms of overall access survival and associated costs [6, 8, 14–20].

All presently approved clinical guidelines recommend performing surveillance of vascular access quality and performance, aiming at early detection of access stenosis, which induced a global trend toward implementation of Qa-based monitoring programs in many dialysis units.

The recommended Qa thresholds for angiography referral are based on its putative predictive power of access malfunction and/or failure.

However, even before the final decision on the clinical relevance of periodic Qa determination, the quality and accuracy of the Qa measurements methods must be questioned, as most techniques have a good correlation among them, but high variation in absolute terms (±200 ml/min) [21–30].

We are now in a position where we feel that we must do some form of VA surveillance, but do not know exactly which. Qa, although not perfect, with results that are hard to interpret and need specific calibration to fine tune appropriate alarm thresholds for each measurement technique, is probably the best hemodynamic parameter to follow.

In our unit, we evaluate monthly Qa, together with a trend analysis of other equally not perfect parameters, like physical examination [31], Kt/V in all dialysis treatments, recirculation, and maximum obtainable Qb with circuit arterial pressure above −250 mmHg, and then decide empirically, as physicians always do, when to refer to angiography.

A successful program of surveillance should reduce thrombosis rate by an amount identical to the angioplasty rate it induces. The key to measure surveillance effectiveness is avoidance of thrombosis; no other surrogate is acceptable.

As a matter of fact, absolute flow (Qa) and drop in flow, measured using several different flow indicators (ultrasound, thermal dilution, ionic dilution), are inaccurate predictors of thrombosis. Most thromboses are unpredictable, and interventions based on surveillance likely yield many unnecessary procedures at high cost.

We do not know if a vascular access defined by us as well functioning actually looks normal in angiography. Without that, it is difficult to really appreciate the specificity of our monitoring indicators and, most of all, the meaning of stenosis in the natural history of the VA.

Our data suggest that the presence of what we call a significant stenosis is not correlated with measured Qa and it might not be associated with early thrombosis deserving immediate intervention [20]. Further studies are needed to clarify

**187**

*Early Detection and Endovascular Intervention to Correct Dialysis Vascular Access Malfunction*

the best surveillance protocol and the role of preemptive intervention in signifi-

**A.** Each unit should perform sequential measurement and trend analysis of the

**B.** Physical examination done and recorded before each dialysis by the R.N., in an access without dressings and needles. Signs to be looked for include a pounding pulse, an intermittent thrill, arm swelling, increment in collateral veins, difficult hemostasis, a new or an enlarging aneurysm, and pain during treatment, reaching an agreement rate with angiogram to detect stenosis of 80% [31].

iv.Native fistulas with a previous Qa ≥ 1000 ml/min—Once a year or whenever

c.Frequent recurrence of significant stenosis (less than 3 months apart).

Patients are referred from the dialysis unit to our VAC by their nephrologists, the indication for intervention is confirmed upon arrival, and an ultrasound/Doppler study will be performed if needed, to decide if it should be referred to the surgical or endovascular arm of the VAC and to help planning the endovascular approach localizing eventual stenotic lesions, their location, and preferred puncture site. **Our referral criteria to surgery**: (a) Native AVF thrombosis; (b) VA rupture; (c) infection with visible abscesses or purulent discharge at puncture sites; (d) need for a new VA; (e) steal syndrome, VA limb distal ischemia; (f) primary malfunction of a VA created or submitted to open surgery less than 1 month ago; (g) growing

**Referral criteria to endovascular intervention:** (a) Growing edema of the VA limb, (b) VA pain during dialysis treatment, (c) recent increment of VA venous pressure associated with a drop in dialysis adequacy, (d) unexplained drop in dialysis adequacy, (e) a drop of VA flow (Qa) in 2 measurements <600 ml/min in a AVG or <400 ml/min a nAVF, (f) need for assisted maturation of a nAVF, (g) superior

The techniques we perform in the angiography suite are (a) diagnostic angiography in no more than 7% of all cases, (b) percutaneous angioplasty (PTA) of stenotic lesions, (c) thrombolysis for thrombosed AVGs, and (d) stenting of elastic

A proposal for surveillance could well include the following:

iii.Native fistulas with a Qa < 1000 ml/min—Quarterly.

a."Last" available vascular access site of that patient.

*DOI: http://dx.doi.org/10.5772/intechopen.92631*

parameters of their choice.

**C.** Access flow measurement (Qa) in:

ii.Other grafts—Quarterly.

clinically indicated.

We consider high-risk grafts:

aneurysm; and (h) hemorrhage.

or frequently relapsed stenosis.

vena cava syndrome, and (h) AVG thrombosis.

b.Frequent clotter.

i.High-risk grafts—Every 2 weeks.

cant stenosis.

*Early Detection and Endovascular Intervention to Correct Dialysis Vascular Access Malfunction DOI: http://dx.doi.org/10.5772/intechopen.92631*

the best surveillance protocol and the role of preemptive intervention in significant stenosis.

A proposal for surveillance could well include the following:

	- i.High-risk grafts—Every 2 weeks.
	- ii.Other grafts—Quarterly.
	- iii.Native fistulas with a Qa < 1000 ml/min—Quarterly.
	- iv.Native fistulas with a previous Qa ≥ 1000 ml/min—Once a year or whenever clinically indicated.

We consider high-risk grafts:


*Cardiac Diseases - Novel Aspects of Cardiac Risk, Cardiorenal Pathology and Cardiac Interventions*

Consensual recommendations for preemptive intervention in malfunctioning grafts are (a) Qa measurement <600 ml/min for grafts or <400 ml/min for native fistulas and (b) a Qa drop higher than 25% over two consecutive measure-

However, recent and quite relevant information has questioned those recommendations, and scanning through recent prospective randomized controlled trials

No matter if we are looking at native fistulas or PTFE grafts, using only Qa measurements, or its association with Doppler studies or dynamic venous pressure as surveillance techniques, it is believed that VA stenosis is now very effectively detected and responsible for a large increase in percutaneous vascular access procedures. Surprisingly, however, it has been found that all these diagnostic and therapeutic procedures fail to reduce the thrombosis rate or prolong access longevity, fueling an ongoing controversy regarding its beneficial effects, both in terms of

All presently approved clinical guidelines recommend performing surveillance of vascular access quality and performance, aiming at early detection of access stenosis, which induced a global trend toward implementation of Qa-based moni-

The recommended Qa thresholds for angiography referral are based on its puta-

However, even before the final decision on the clinical relevance of periodic Qa determination, the quality and accuracy of the Qa measurements methods must be questioned, as most techniques have a good correlation among them, but high

We are now in a position where we feel that we must do some form of VA surveillance, but do not know exactly which. Qa, although not perfect, with results that are hard to interpret and need specific calibration to fine tune appropriate alarm thresholds for each measurement technique, is probably the best hemodynamic

In our unit, we evaluate monthly Qa, together with a trend analysis of other equally not perfect parameters, like physical examination [31], Kt/V in all dialysis treatments, recirculation, and maximum obtainable Qb with circuit arterial pressure above −250 mmHg, and then decide empirically, as physicians always do, when

A successful program of surveillance should reduce thrombosis rate by an amount identical to the angioplasty rate it induces. The key to measure surveillance

We do not know if a vascular access defined by us as well functioning actually looks normal in angiography. Without that, it is difficult to really appreciate the specificity of our monitoring indicators and, most of all, the meaning of stenosis in

Our data suggest that the presence of what we call a significant stenosis is not correlated with measured Qa and it might not be associated with early thrombosis deserving immediate intervention [20]. Further studies are needed to clarify

effectiveness is avoidance of thrombosis; no other surrogate is acceptable. As a matter of fact, absolute flow (Qa) and drop in flow, measured using several different flow indicators (ultrasound, thermal dilution, ionic dilution), are inaccurate predictors of thrombosis. Most thromboses are unpredictable, and interventions based on surveillance likely yield many unnecessary procedures at

measurement," and "Monitoring plus intervention reduces thrombotic rates,

morbidity and costs" [11, 12].

in this field reveals some discordant opinions.

toring programs in many dialysis units.

parameter to follow.

to refer to angiography.

the natural history of the VA.

overall access survival and associated costs [6, 8, 14–20].

tive predictive power of access malfunction and/or failure.

variation in absolute terms (±200 ml/min) [21–30].

ments [13].

**186**

high cost.

c.Frequent recurrence of significant stenosis (less than 3 months apart).

Patients are referred from the dialysis unit to our VAC by their nephrologists, the indication for intervention is confirmed upon arrival, and an ultrasound/Doppler study will be performed if needed, to decide if it should be referred to the surgical or endovascular arm of the VAC and to help planning the endovascular approach localizing eventual stenotic lesions, their location, and preferred puncture site.

**Our referral criteria to surgery**: (a) Native AVF thrombosis; (b) VA rupture; (c) infection with visible abscesses or purulent discharge at puncture sites; (d) need for a new VA; (e) steal syndrome, VA limb distal ischemia; (f) primary malfunction of a VA created or submitted to open surgery less than 1 month ago; (g) growing aneurysm; and (h) hemorrhage.

**Referral criteria to endovascular intervention:** (a) Growing edema of the VA limb, (b) VA pain during dialysis treatment, (c) recent increment of VA venous pressure associated with a drop in dialysis adequacy, (d) unexplained drop in dialysis adequacy, (e) a drop of VA flow (Qa) in 2 measurements <600 ml/min in a AVG or <400 ml/min a nAVF, (f) need for assisted maturation of a nAVF, (g) superior vena cava syndrome, and (h) AVG thrombosis.

The techniques we perform in the angiography suite are (a) diagnostic angiography in no more than 7% of all cases, (b) percutaneous angioplasty (PTA) of stenotic lesions, (c) thrombolysis for thrombosed AVGs, and (d) stenting of elastic or frequently relapsed stenosis.

In our unit, prospective results of 1-year follow-up in 71 new AV grafts with monthly surveillance revealed the following:


e.A thrombosis rate—0.46 thru/pt. year.

f. In 60% of cases, previous monitoring was normal.
