Anda di halaman 1dari 8

Clin Kidney J (2015) 8: 282289

doi: 10.1093/ckj/sfv019
Advance Access publication 2 April 2015

CKJ review

Current tools for prediction of arteriovenous stula outcomes

Damian G. McGrogan1, Alexander P. Maxwell2,3, Aurang Z. Khawaja1 and Nicholas G. Inston1

1
Department of Vascular Access and Renal Transplantation, University Hospitals Birmingham, Queen Elizabeth Hospital, Birmingham
B15 2TH, UK, 2Regional Nephrology Unit, Belfast Health and Social Care Trust, Belfast City Hospital, Belfast BT9 7AB, UK and 3School of
Medicine, Dentistry and Biomedical Sciences, Health Sciences Building, Belfast BT9 7BL, UK
Correspondence to: Damian G. McGrogan; E-mail: damianmcg@gmail.com; damian.mcgrogan@uhb.nhs.uk

Abstract
It remains challenging to accurately predict whether an individual arteriovenous stula (AVF) will
mature and be useable for haemodialysis vascular access. Current best practice involves the use of
routine clinical assessment and ultrasonography complemented by selective venography and
magnetic resonance imaging. The purpose of this literature review is to describe current practices
in relation to pre-operative assessment prior to AVF formation and highlight potential areas for
future research to improve the clinical prediction of AVF outcomes.

Keywords: arteriovenous stula; clinical assessment; denitions; ultrasonography; venography

Introduction highlight potential areas for future research to improve


the clinical prediction of AVF outcomes.
It remains challenging to accurately predict whether an
individual arteriovenous stula (AVF) will mature and be
useable for haemodialysis vascular access. In part, this is
due to the heterogeneity of end-stage renal disease How an AVF matures
(ESRD) populations studied as these have varied in terms
of age structure, background ethnicity and the relative Guidelines suggest attempting to create an AVF at the
prevalence of different ESRD aetiologies. The natural most appropriate distal arm site. In practice, this means
history of AVF maturation is also confounded by multiple choosing the more proximal radiocephalic site before con-
comorbid conditions, such as diabetes and peripheral sidering a brachiocephalic AVF. If these sites are unsuit-
vascular disease, which may be present in ESRD patients. able then a brachiobasilic procedure or placement of an
The outcome (clinically usable AVF) is further impacted by arteriovenous graft (AVG) may be most appropriate [1].
variation in point-of-care management of AVFs, e.g. can- A useable AVF can offer the haemodialysis patient a
nulation expertise. In order to study the effect any variable durable vascular access option which, in comparison to
has on a given vascular access outcome, it is important to prolonged CVC dependence, is associated with a lower risk
establish specic denitions. The fact that there are mul- of bloodstream infection and central venous stenosis [2].
tiple denitions of vascular access outcomes (summarized The ideal AVF must achieve an adequate connection
in Table 1) makes it more difcult to compare published between the high-pressure arterial and low-pressure
data from different centres and countries. venous systems; bypassing the small capillary network of
The Renal Association Vascular Access for Haemodialy- the palmar arch while maintaining arterial perfusion and
sis clinical practice guidelines [1] have encouraged the venous drainage in the tissues distal from the site of cre-
earlier formation of AVFs in an attempt to reduce the ation [3]. As a result of the arteriovenous anastomosis,
number of ESRD patients commencing haemodialysis pressure within the vein increases, causing dilatation and
using a central venous catheter (CVC). Furthermore, in the arterialization of the anastomosed vein. A reex increase
UK, National Health Service best practice tariffs for in cardiac output allows for this distal tissue perfusion and
haemodialysis promote the use of AVFs as renal units pro- maturation of the stula which is frequently complemen-
viding haemodialysis receive higher annual payments for ted by retrograde arterial ow [4]. Where retrograde ow
those ESRD patients using an AVF (compared with patients is excessive, steal syndrome may occur characterized by
with a CVC). These nancial incentives coupled with the distal neuralgia and eventually peripheral limb ischaemia
persistently high failure rate of AVFs support renewed [5, 6].
efforts to identify pre-operative predictors of AVF out- Poiseuilles law [as shown in Equation (1)] describes
comes. The purpose of this literature review is to describe essential determinants of blood ow which are related to
current practices in relation to pre-operative assessment perfusion pressure, blood vessel radius and length and the
prior to AVF formation as summarized in Figure 1 and viscosity of blood. The maturation of an AVF is partly

The Author 2015. Published by Oxford University Press on behalf of ERA-EDTA.This is an Open Access article distributed under the terms of the Creative
Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and
reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017
Current tools for prediction of AVF outcomes 283
Table 1. Denitions of stula outcomes

Primary failure Immediate failure of AVF within 72 h of surgery, early dialysis suitability failure or late dialysis suitability failure
(NAVAC denition)
Early dialysis This is an AVF for which, despite interventions (radiologic or surgical), it was not possible
suitability failure to use the AVF successfully for haemodialysis by the third month following its creation
(NAVAC denition)
Late dialysis This is an access for which, despite interventions (radiologic or surgical), it was not
suitability failure possible to use the AVF successfully for haemodialysis by the sixth month following its
creation (NAVAC denition)
Primary patency The interval from the time of access creation until the rst access thrombosis or any intervention to maintain or
restore blood ow (NAVAC and SVS denition)
Assisted primary patency The interval from time of access placement to access thrombosis or time of measurement of patency, including
(thrombosis-free access survival) intervening manipulations (surgical or endovascular interventions) designed to maintain the functionality of a
patent access (not recommended by NAVAC or SVS)
Functional primary patency The time from the rst successful two-needle cannulation until the rst intervention or access failure
Secondary (cumulative) patency The time from access creation until access abandonment. Secondary patency was not terminated by surgical or
interventional radiology procedures to maintain or restore patency (NAVAC and SVS denition)
Functional secondary patency The interval from the rst successful two-needle cannulation for haemodialysis treatment to access
abandonment (NAVAC and SVS denition)

AVF, arteriovenous stula; NAVAC, North American Vascular Access Consortium; SVS, Society for Vascular Surgery.

governed by these parameters. antiplatelet agents in chronic kidney disease found that
while there was an associated reduction in early thrombosis
Pressure difference  Radius4 rates of AVF, this had no effect on suitability for dialysis sup-
Volume flow rate 1 porting other underlying reasons [13]. For the purposes of
8=p  Viscosity  Length pre-operative evaluation, the presence of such factors may
be important but are generally not modiable or useful in in-
Normal brachial arterial blood ow is 50 mL/min with dividuals to plan access.
radial blood ow rates of 25 mL/min [7]. This blood ow
rate has been shown to increase 3- to 5-fold during exer-
cise or following reactive hyperaemia; however, according Pre-operative assessment
to Poiseuilles law, radial artery ow must increase 20-fold
in order to deliver >500 mL/min of blood ow and achieve
Clinical assessment
radiocephalic AVF maturation. The actual measured
increase in blood ow is only 4050% of this prediction Clinical assessment of an ESRD patients suitability for AVF
and therefore other mechanisms must be operating [8]. creation involves detailed history taking and appropriate
The increase in blood ow is achieved through increased physical examination as summarized in Table 2. The use of
antegrade arterial blood ow in addition to retrograde clinical assessment alone may provide sufcient assess-
blood ow from the palmar arterial arch in 75% of AVFs ment in selected cases. Wells et al. [14] discuss the select-
[4], decreased blood viscosity associated with the increase ive use of ultrasonographic vascular mapping in the
in ow rate and therefore reduction in wall shear stress evaluation of patients prior to the start of haemodialysis.
and friction [8]. This occurs in combination with main- They indicate that for some patients, ultrasound vascular
tained antegrade ow throughout the cardiac cycle as mapping is unnecessary and allows patients to avoid any
opposed to retrograde ow seen in diastole in a normal time delay associated with an extra clinic appointment for
artery [9]. ultrasonography and also reduces overall costs associated
Other processes are involved in AVF maturation as with preoperative assessment. They suggest that this facil-
shown in Figure 2. Arterial disease in the form of arterio- itates the development of a one stop shop whereby
sclerosis and pre-existing neointimal thickening asso- patients undergo clinical assessment and surgery on the
ciated with conditions such as diabetes, hypertension and same day. It could also be argued that training of vascular
ESRD itself may contribute to the development of neointi- access surgeons in the use of ultrasonographic techniques
mal hyperplasia after AVF formation [10]. Endothelial dys- specic to vascular access would also reduce the costs
function can also impair the arterial response to increased incurred and time delays.
wall shear stress, therefore inhibiting the dilatation and in-
creased ow rates required for AVF maturation [3]. Venous
Patient and stula selection
trauma and brosis caused by previous cannulation, vene-
puncture and the surgical creation of an AVF can lead to Age and gender are often natural discriminators. As with
inability of venous distension preventing ow-induced re- many conditions, the average age of the population with
modelling [11]. Pre-existing arterial or venous disease and ESRD is increasing annually. Hod et al. [15] highlighted
the surgical procedure itself are important components of that the important risk factors for AVF failure in those
the success or failure of an AVF. older than 67 years of age are increasing age, female
The effect of coagulation disorders (high factor VIII:C, gender, black race, diabetes, cardiac failure and a shorter
homocysteine, brinogen, d-dimer; presence of antiphospho- period of pre-ESRD nephrology care. Similarly, Miller et al.
lipid antibodies and short thrombin time) and thrombophilia [16] have shown a lower AVF adequacy in women com-
(factor V Lieden and prothrombin G20210A mutations; pared with men which was unexplained by pre-operative
protein C and antithrombin activities; and protein S) on mat- blood vessel diameters. Ng et al. [17] found from their
uration is unknown, although female sex and thrombophilia analysis of 2920 patients that AVGs, female gender, dia-
are independent risk factors for loss of primary patency [12]. betes and advanced age were all associated with signi-
A recent Cochrane systematic review on the use of cantly shorter primary AVF patency.

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017
284 D.G. McGrogan et al.

Fig. 1. Summary of advantages and disadvantages associated with preoperative assessment tools. DSA, digital subtraction angiography; MRV, magnetic
resonance venography; DUS, Doppler ultrasound scan; MRI, magnetic resonance imaging; NSF, nephrogenic systemic brosis; ESRD, end-stage renal
disease; AGE, advanced glycation end-products.

In a systematic review performed by Al-Jaishi et al., it patency rates at 1 year were signicantly worse in lower
was shown that the primary AVF failure rates were higher arm versus upper arm sites (55 versus 65% for lower and
in more distal compared with proximal upper limb sites upper arm AVFs, respectively) [18].
[ primary failure rates of 28% for lower arm AVF and 20% Older age, female gender, presence of diabetes and
for upper arm AVF (P = 0.001)]. Similarly, the primary AVF distal AVFs have all been identied as risk factors for

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017
Current tools for prediction of AVF outcomes 285

Fig. 2. Summary of factors involved in AVF maturation.

Table 2. Summary of clinical assessment prior to arteriovenous stula mm for arterial diameter as they found this to be asso-
formation ciated with a very low early AVF failure rate of 8% and an
excellent 1-year AVF patency rate of 83%. AVF patency
Clinical assessment for vascular access
History Previous central venous beyond 6 weeks is considered an important parameter in
catheterization the analysis of AVF outcomes.
Presence of diabetes and DUS can be used to assess the functional ability of the
hypertension artery to dilate by using a reactive hyperaemia test. This is
Use of anticoagulant or antiplatelet
agents performed by asking the patient to clench their st while
Peripheral arterial assessment Palpation of axillary, brachial, radial observing the phasic ow via DUS. The patient then re-
and ulnar pulses bilaterally noting leases the st and the arterial reaction is measured. This
presence, absence or diminished simulates the change from high pressure to low pressure
character
Bilateral blood pressure assessing for which occurs in AVF formation and therefore provides a
discrepancies surrogate measurement of how the artery will respond to
Allens test the stress of AVF surgery. Malovrh [19] demonstrated that
Peripheral venous assessment Assess cephalic and basilic systems an absence of reactive hyperaemia characterized by a re-
(performed with proximal Assess for patency, presence of a
tourniquet) linear segment, collateral sistance index of <0.7 on opening of the st is predictive of
engorgement of chest wall veins immediate postoperative AVF failure. These ndings have
been challenged by Lockhart et al. [25] who found no dif-
ference in AVF outcomes, except in functional patency of
primary AVF failure and should be taken into consideration AVFs in women and Wall et al. [26] who found no differ-
when planning an AVF procedure. ence in primary functional AVF patency but a signicant
difference in secondary AVF patency following surgical re-
vision. These studies do not lend themselves to direct
Doppler ultrasound
comparison but do suggest that the ability of the artery to
Doppler ultrasound (DUS) is one of the key techniques dilate plays a signicant functional role in AVF maturation.
currently employed for enhancing clinical examination of DUS is also used to assess the cephalic and basilic veins
the patient prior to AVF construction. This is used not only in the arm, including the venous wall, route, patency,
as a preoperative tool in assessing arterial and venous calibre, distensability and collateral circuits. The ideal vein
anatomy but also in post-operative monitoring of AVF mat- for AVF formation has a relatively straight course and
uration and ongoing AVF surveillance. DUS assessment should lie <6 mm from the skin surface [27].
does require more clinical skill, equipment and time to Due to the supercial nature of the veins, transducer gel
perform (compared with clinical examination alone), but it should be applied copiously to prevent excessive pressure
remains non-invasive, safe, has reproducible results and on the vein in question and therefore underestimation of
helps to identify clinically occult veins [19]. DUS allows for vein diameter. Several studies have suggested the ideal
observation of arterial diameter, vessel wall thickness, wall diameter of the vein to be used for AVF formation. Small
alterations, blood vessel course, localization of any ob- veins of <1.6 mm in diameter are associated with a high
structive or stenotic lesions present and can also perform a risk of early AVF failure within 12 weeks [23]. Silva et al.
functional assessment. The 2007 European Renal Best Prac- [24] suggested a minimum diameter of 2.5 mm with a
tice guidelines suggest the use of DUS assessment in all pa- tourniquet applied, whereas Mendes et al. [28] suggest a
tients being considered for vascular access formation [20]. minimum diameter of >2 mm.
A number of studies have been conducted to assess the Venous distensability is assessed by measuring the
utility of arterial diameter in predicting AVF outcomes. diameter of the vein before and after at least 2 min of
Malovrh [21] reported immediate AVF failure rates of 44% tourniquet placement. This can be achieved by inating a
and early failure rates of 64% where radial arterial dia- sphygmomanometer to 60 mmHg. The percentage in-
meters were <1.5 mm on pre-operative DUS. This con- crease in the size of the vein is then evaluated by ultra-
trasts with immediate and early AVF failure rates of 8 and sound. Malorvh [29] found that venous distensability was
17%, respectively, where radial arterial diameters were predictive of outcomes since subsequently successful AVFs
>1.5 mm [22]. These ndings have been supported by showed a mean percentage dilatation of 44% compared
Parmar et al. [22] who found arterial diameters of <1.5 with only 11% in the unsuccessful AVF group. Lockhart
mm had an AVF failure rate of 45% compared with 0% et al. [30] similarly found DUS useful in the identication
when radial arterial diameter was >1.5 mm. Wong et al. of suitable veins by concluding that veins with a luminal
[23] found a signicant difference at an arterial diameter diameter of >2.5 mm and those smaller veins that dilated
cut-off of 1.6 mm. Silva et al. [24] proposed a cut-off of 2 up to 2.5 mm with placement of a tourniquet were equally

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017
286 D.G. McGrogan et al.
suitable for AVF formation. The presence of accessory Magnetic resonance imaging
veins has also been suggested to be a factor in non-mat-
Magnetic resonance angiography (MRA) has been shown
uration of AVF. An accessory vein <5 cm from the anasto-
to enable accurate detection of upper limb arterial and
mosis can alter the functionality of the AVF [23], while the
venous stenosis and occlusions prior to AVF creation [41].
size of the collaterals rather than their position may
These stenoses are undetectable by conventional DUS
impact on the rate of AVF non-maturation [31].
and therefore identication could potentially reduce the
A recent systematic review [32] has concluded that the
early AVF failure and non-maturation rate. The recognition
use of pre-operative DUS is associated with a higher rate of
of nephrogenic systemic brosis and its association with
commencing dialysis, but this association did not reach
gadolinium-based contrast media has prompted compari-
statistical signicance and therefore additional novel pre-
son of contrast-enhanced versus non-contrast-enhanced
dictors of AVF outcomes may help improve clinical manage-
MRA [42] (NCE-MRA) and shown that while image quality
ment of ESRD patients. This systematic review conclusion
and vessel-to-background ratios were lower, NCE-MRA is a
is supported by Smith et al. [33] who randomized patients
feasible alternative in patients with ESRD.
into either routine or selective preoperative ultrasound
Merkx et al. [43] assessed the merits of NCE-MRA for pre-
imaging. They reported that routine pre-operative does not
dicting vascular access outcomes based around a com-
reduce early failure rates, inuence site of AVF formation or
puter model. Postoperative arm inows were predicted by
reduce complications concluding that routine preoperative
computer modelling following pre-operative NCE-MRA and
imaging may not be necessary where clinical evaluation
compared with blood ow measurements assessed by DUS
detects suitable anatomy for AVF formation.
postoperatively. They concluded that NCE-MRA is able to
A novel computational model has been suggested by
provide geometrical details of arterial stenoses which could
Caroli et al. [34] which is completely automated, fast, in-
assist the vascular surgeon in AVF planning. This expensive
volves operator-independent calculations and enables the
and less readily available investigation is unlikely to replace
observer to quantitatively estimate patient-specic post-
clinic-based DUS in the assessment of patients for AVF for-
operative blood ow volume change over different AVF
mation but may have applications in patients with periph-
congurations. This model uses preoperative vessel di-
eral arterial disease, thrombosis or reduced arm inows
mension and arterial blood ow volume measurements
due to local stenoses or global narrowing within the arterial
taken during ultrasound assessment as input parameters
lumen. Another limitation of this method is the necessity
to predict postoperative diameters and blood ow volumes
for DUS assessment of venous diameters since NCE-MRA
at different time points after surgery. The computational
grossly overestimates venous diameters [44].
model was found to be accurate in predicting blood ow
volumes in individual patients 40 days postoperatively
with highly signicant correlation for different AVFs. As
Roy-Chaudhury et al. [35] comment, this is a step towards Novel techniques for prediction of AVF outcomes
individualization of AVF creation and a good example of
the multidisciplinary research collaboration necessary to Assessment of arterial stiffness
achieve a signicant impact on vascular access outcomes.
Arterial stiffness, which is common in patients with chronic
kidney disease, refers to the distensability, compliance and
elastic modulus of the arterial vascular system and can be
measured locally, regionally and systemically. Aortic pulse
Venography wave velocity is considered the gold standard [45] for
Digital subtraction venography is considered to be the gold assessing arterial stiffness as it gives the clearest patho-
standard for assessment of the venous system and has physiological signicance since the majority of the buffer-
been investigated as a potential adjunct to pre-operative ing of pulse waves occurs within the aorta. The pulse wave
assessment of patients with ESRD referred for AVF forma- Vicorder (Smart Medical, Moreton-in-Marsh, UK) is a non-
tion. Traditional iodinated contrast can be used in patients invasive, easy to learn and reproducible method of asses-
who are being dialysed; however, in patients who are pre- sing stiffness along an arterial section. Although other
dialysis, magnetic resonance venography had been shown systems exist to measure aortic pulse wave velocity such
to have acceptable sensitivity and specicity when com- as the SphygmoCor system (AtCor Medical, West Ryde,
pared with venography with good inter-observer correlation NSW, Australia) and Complior apparatus (Artech Medical,
regarding imaging quality and strategy planning [36]. Pantin, France), these require greater operator instruction
Gadolinium-based magnetic resonance contrast media are and are more intrusive to patients as they require palpation
no longer considered safe for persons with chronic kidney and assessment at the femoral artery. Increased aortic
disease because of the risk of nephrogenic systemic brosis pulse wave velocity has been independently associated
[37]. Carbon dioxide-based venography is a promising alter- with adverse cardiovascular outcomes in large prospective
native contrast choice for use in patients with advanced studies, including specically patients with ESRD [46]. As-
chronic kidney disease and ESRD [38]. sessment of arterial stiffness has yet to be extensively in-
Hyland et al. [39] have shown that venography is able to vestigated in relation to AVF outcomes and is an area of
identify clinically occult veins which may be usable for AVF interest for future research projects.
formation, a conclusion which is supported by Patel et al.
[40] who found that while this imaging modality can iden-
Assessment of endothelial function
tify a greater number of suitable veins, paradoxically, the
combined use of DUS and venography was associated with The responsiveness of the endothelium is an important
a decreased AVF maturation rate. The 2007 European Best aspect of the remodelling required to establish a mature
Practice Guidelines advocate central venous imaging for all AVF. The endothelium acts as an interface between the
patients with a history of CVC placement to identify central blood and all other tissues in the body performing import-
venous stenosis or occlusion prior to AVF formation [21]. ant roles in maintaining the vascular environment through

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017
Current tools for prediction of AVF outcomes 287
release of agents that regulate vasomotor tone, inamma- identied as having prognostic value in predicting mortal-
tory responses and homeostatic functions [46]. Nitric oxide ity in patient with ESRD-related bone mineral disease
is a potent vasodilator, inhibitor of inammatory activity, while elevated C-reactive protein, interleukin-6 and
smooth muscle cell proliferation and platelet adhesion and tumour necrosis factor levels are associated with in-
aggregation [47]. The loss of the vasodilatory response and creased mortality in ESRD patients [54]. Asymmetric di-
promotion of thrombosis, inammation and cellular prolif- methylarginine is an endogenous inhibitor of nitric oxide
eration associated with endothelial dysfunction is a poten- synthase and high circulating levels are associated with
tially worthwhile research topic, since these mechanisms endothelial dysfunction and atherosclerosis in patients
are closely related to AVF maturation. A number of studies with pre-dialysis ESRD [55].
have linked endothelial dysfunction with uraemia and A number of inammatory markers, such as interleukin-
reduced nitric oxide production in chronic renal disease 6, tumour necrosis factor- and soluble P selectin, have
[48]. Owens et al. [49] found evidence that impairment of been associated with adverse cardiovascular outcomes
endothelial function is associated with decreased arterial [56]; however, high-sensitivity CRP (hs-CRP) has the stron-
remodelling and nal venous diameter attained at 3 gest association with cardiovascular outcomes [57]. It is
months post-stula formation. reported that hs-CRP can destabilize nitric oxide synthase
Two methods of measuring endothelial dysfunction are mRNA in endothelial cells [58], suggesting a correlation
ow-mediated dilatation (FMD) and peripheral arterial to- between high circulating levels of hs-CRP and endothelial
nometry (PAT). FMD is the most well-established method dysfunction. Also, 13-hydroxyoctadecadienoic acid (13-
of assessing endothelial function and is based on the HODE) has been found to decrease with age in patients
reactive hyperaemic response following a period of occlu- undergoing coronary artery bypass grafting and has been
sion with a tourniquet. Increased blood ow following associated with an increase in endothelial cell thrombo-
release of the cuff causes increased wall shear stress, genicity [56]. Other surrogate markers of endothelial cell
leading to release of nitric oxide and therefore vasodilata- activation and inammation have been proposed, including
tion [50]. Monitoring the response will give a surrogate lipoprotein-associated phospholipase A2, CD40 receptor/
marker of endothelial performance post-stula formation CD40 ligand interaction, LOX-1 and direct measurements of
and therefore guide vascular surgeons regarding patient circulating endothelial progenitor cells [59]. Further studies
suitability and vascular access site placement. PAT is a are required to establish if assessing biomarkers of baseline
more recent development which uses pneumatic nger endothelial function in patients with ESRD is clinically
probes to measure the digital pulse wave amplitude when useful either for prediction of AVF maturation or to target
reactive hyperaemia is induced [51]. A recent study asses- therapies to modify local vascular biology prior to AVF
sing the correlation between these two methods in formation.
healthy patients and patients with peripheral arterial Creation of an AVF has been shown to have signicant
disease concluded that there was no correlation between effects on atrial natriuretic peptide (ANP) and brain natri-
these two methods and that FMD is a more accurate uretic peptide (BNP). An increase in ANP was induced by
measure of nitric oxide-induced endothelial function [52]. volume loading and BNP release was stimulated by left
A novel method of assessing the hyperaemic response ventricular diastolic dysfunction. Additionally, a positive
in patients undergoing AVF formation is the use of near-in- relationship has been shown between MMP-2 gene ex-
frared spectroscopy. This is a non-invasive, continuous, pression in vein segments evaluated by gelatin zymogra-
real-time determination of chromophore concentration phy and western blotting and AVF maturation [60]. The
on the basis of spectrophotometric principles. In mamma- maximal benet from biomarkers may lie in the multi-
lian tissue, there are three chromophores which demon- marker approach as evidenced by the Olmsted study
strate absorption spectra in the near-infrared range (800 which reported that the combination of increased BNP
1200 nm wavelength), namely haemoglobin and deoxy- with CRP levels enhanced the ability to predict risk of
haemoglobin, myoglobin and cytochrome oxidase. The death compared with standard risk factors for mortality in
result is a measurement of mixed arterial and venous heart failure [61].
oxygen concentrations within the tissue being studied Advanced glycation end-products (AGEs) have been im-
which, if placed distal to the site of AVF planning, may plicated in the pathophysiology of vascular disease where
provide new information about the physiological response they accumulate in the endothelial cell wall and disrupt
to AVF formation and assist in AVF planning. Another alter- cellular structure and function, including decreasing the
native way to assess endothelial dysfunction is the use of bioavailability of nitrous oxide and alteration of cell
laser Doppler owmetry and imaging to measure cutane- surface structure from an anticoagulant to pro-coagulant
ous perfusion accompanied by iontophoresis of acetylcho- state. In theory, reducing the burden of AGEs would offer
line and sodium nitroprusside. Hansell et al. [53] found a a potential novel therapeutic option to improve vascular
signicant correlation between FMD in the brachial artery biology and potentially improve maturation of AVFs. There
and laser Doppler owmetry in small cutaneous vessels is experimental evidence that agents such as aminogua-
supporting the systemic nature of endothelial dysfunction nidine, ALT-946, ALT-711, statins, pyridoxamine and
and that microvascular changes in the skin mimic macro- dietary modications can help to reduce AGE levels [62].
vascular changes in the supplying artery. There is no convincing evidence yet that such therapies
improve vascular access outcomes.
Biomarkers
Assisted maturation
As yet, there have been no biomarkers identied as a
useful adjunct to AVF planning. Biomarkers have been ex- Several novel therapies have been proposed to assist AVF
tensively investigated in ESRD patients focusing on their maturation, including the use of far-infrared (FIR)
usefulness in prognosis, diagnostics, response to therapy therapy [63]. FIR is a non-invasive therapeutic modality
and prevention of disease. Elevated broblast growth which has been reported to improve access ow rates
factor 23 and bone alkaline phosphatase have been and reduce the incidence of AVF malfunction in

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017
288 D.G. McGrogan et al.
haemodialysis patients. Primary balloon angioplasty has 7. Joannides R, Bakkali EH, Le Roy F et al. Altered ow-depend-
been shown to improve primary AVF patency rates for ent vasodilatation of conduit arteries in maintenance haemo-
AVFs with venous diameters of <3 mm [64] and also in dialysis. Nephrol Dial Transplant 1997; 12: 26232628
AVFs with venous diameters of 2 mm [65]. These con- 8. Dammers R, Tordoir JH, Kooman JP et al. The effect of ow
clusions have been supported by more recent data from changes on the arterial system proximal to an arteriovenous
Turmel-Rodrigues, suggesting that by using an aggres- stula for hemodialysis. Ultrasound Med Biol 2005; 31:
sive, multi-disciplinary approach a non-maturing AVF can 13271333
9. Remuzzi A, Ene-Iordache B, Mosconi L et al. Radial artery wall
be identied, evaluated and salvaged [66]. An increased
shear stress evaluation in patients with arteriovenous stula
number of interventions required to assist maturation is
for hemodialysis access. Biorheology 2003; 40: 423430
however associated with reduced primary and secondary 10. Chowdhury UK, Airan B, Mishra PK et al. Histopathology and
patency rates and a higher number of interventions to morphometry of radial artery conduits: basic study and clinic-
maintain patency [67]. al application. Ann Thorac Surg 2004; 78: 16141621
11. Konner K, Nonnast-Daniel B, Ritz E. The arteriovenous stula.
J Am Soc Nephrol 2003; 14: 16691680
Current studies 12. Salmela B, Hartman J, Peltonen S et al. Thrombophilia and ar-
The Hemodialysis Fistula Maturation [68] study is currently teriovenous stula survival in ESRD. Clin J Am Soc Nephrol
2013; 8: 962968
underway in the USA. This is a multicentre prospective
13. Palmer SC, Di Micco L, Razavian M et al. Antiplatelet agents
cohort study which is aiming to recruit 600 patients under-
for chronic kidney disease. Cochrane Database Syst Rev 2013;
going new AVF creation with a 4-year follow-up period. 2: CD008834
The primary outcome is unassisted clinical maturation. The 14. Wells AC, Fernando B, Butler A et al. Selective use of ultra-
results of this study are eagerly anticipated as this is the sonographic vascular mapping in the assessment of patients
rst, large-scale, adequately powered cohort to address before haemodialysis access surgery. Br J Surg 2005; 92:
many of the outstanding questions related to predictors of 14391443
AVF outcomes. Further studies into vascular access out- 15. Hod T, Desilva RN, Patibandla BK et al. Factors predicting
comes include the Antiaggregation in Primary Prevention failure of AV stula rst policy in the elderly. Hemodial Int
of Vascular Access for Hemodialysis (NCT02055131) and 2014; 18: 507515
The Use of Glyceryl Trinitrate Patches in Arteriovenous Fis- 16. Miller CD, Robbin ML, Allon M. Gender differences in outcomes
tulas (NCT01685710) which will provide further informa- of arteriovenous stulas in hemodialysis patients. Kidney Int
tion on the use of aspirin in varying doses and the use of 2003; 63: 346352
glyceryl trinitrate patches applied 5 cm proximal from the 17. Ng YY, Wu SC, Hung YN et al. Effect of demographic character-
site of vascular access surgery in the prevention of early istics and timing of vascular access maturation on patency in
stenoses. Chinese incident haemodialysis patients. Nephrol Dial Trans-
plant 2009; 24: 34473453
18. Al-Jaishi AA, Oliver MJ, Thomas SM et al. Patency rates of the
arteriovenous stula for hemodialysis: a systematic review
Conclusion and meta-analysis. Am J Kidney Dis 2014; 63: 464478
19. Malovrh M. Native arteriovenous stula: preoperative evalu-
The factors leading to failure of AVF maturation are still ill ation. Am J Kidney Dis 2002; 39: 12181225
dened and there is a limited ability to predict surgical 20. Tordoir J, Canaud B, Haage P et al. EBPG on vascular access.
vascular outcomes. Clinical examination and DUS mea- Nephrol Dial Transpl 2007; 22(Suppl 2): ii88ii117
21. Malovrh M. Non-invasive evaluation of vessels by duplex son-
surements are the mainstays of current preoperative as-
ography prior to construction of arteriovenous stulas for
sessment. Further studies into this under-researched eld
haemodialysis. Nephrol Dial Transplant 1998; 13: 125129
are currently underway and there are many opportunities 22. Parmar J, Aslam M, Standeld N. Pre-operative radial arterial
to develop better tools to predict AVF outcomes. diameter predicts early failure of arteriovenous stula (AVF) for
haemodialysis. Eur J Vasc Endovasc Surg 2007; 33: 113115
Conict of interest statement. None declared.
23. Wong V, Ward R, Taylor J et al. Factors associated with early
failure of arteriovenous stulae for haemodialysis access. Eur
J Vasc Endovasc Surg 1996; 12: 207213
REFERENCES 24. Silva MB Jr, Hobson RW 2nd, Pappas PJ et al. A strategy for in-
creasing use of autogenous hemodialysis access procedures:
1. Vascular Access Work Group. Clinical practice guidelines for
impact of preoperative non invasive evaluation. J Vasc Surg
vascular access. Am J Kidney Dis 2006; 48 (Suppl 1):
1998; 27: 302307
S248S273
25. Lockhart ME, Robbin ML, Allon M. Preoperative sonographic
2. Fysaraki M, Samonis G, Valachis A et al. Incidence, clinical,
radial artery evaluation and correlation with subsequent
microbiological features and outcome of bloodstream infec-
radiocephalic stula outcome. J Ultrasound Med 2004; 23:
tions in patients undergoing hemodialysis. Int J Med Sci 2013;
161168
10: 16321638
26. Wall LP, Gasparis A, Callahan S et al. Impaired hyperemic re-
3. Dixon BS. Why dont stulas mature? Kidney Int 2006; 70:
sponse is predictive of early access failure. Ann Vasc Surg
14131422
2004; 18: 167171
4. Sivanesan S, How TV, Bakran A. Characterizing ow distribu-
27. NKF-K/DOQI. Update vascular access. Guideline 1: patient
tions in AV stulae for haemodialysis access. Nephrol Dial
preparation for permanent hemodialysis access. Am J Kidney
Transplant 1998; 13: 31083110
Dis 2006; 48(Suppl 1): s188s191
5. Billet A, Queral LA, Polito WF et al. The vascular steal phenom-
28. Mendes RR, Farber MA, Marston WA et al. Prediction of wrist ar-
enon: an experimental model. Surgery 1984; 96: 923928
teriovenous stula maturation with preoperative vein mapping
6. Ramuzat A, How TV, Bakran A. Steal phenomenon in radioce-
with ultrasonography. J Vasc Surg 2002; 36: 460463
phalic arteriovenous stula. In vitro haemodynamic and elec-
29. Malovrh M. The role of sonography in the planning of arterio-
trical resistance simulation studies. Eur J Vasc Endovasc Surg
venous stulas for hemodialysis. Semin Dial 2003; 16: 299303
2003; 25: 246253

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017
Current tools for prediction of AVF outcomes 289
30. Lockhart ME, Robbin ML, Fineberg NS et al. Cephalic vein stage IV and V chronic kidney disease. J Vasc Access 2010; 11:
measurement before forearm stula creation: does use of a 329334
tourniquet to meet the venous diameter threshold increase 50. Lekakis J, Abraham P, Balbarini A et al. Methods for evaluating
the number of usable stulas? J Ultrasound Med 2006; 25: endothelial function: a position statement from the European
15411545 Society of Cardiology Working Group on Peripheral Circulation.
31. Beathard GA, Arnold P, Jackson J et al. Aggressive treatment Eur J Cardiovasc Prev Rehabil 2011; 18: 775789
of early stula failure. Kidney Int 2003; 64: 14871494 51. Kuvin JT, Patel AR, Sliney KA et al. Assessment of peripheral
32. Wong CS, McNicholas N, Healy D et al. A systematic review of vascular endothelial function with nger arterial pulse wave
preoperative duplex ultrasonography and arteriovenous amplitude. Am Heart J 2003; 146: 168174
stula formation. J Vasc Surg 2013; 57: 11291133 52. Allan RB, Delaney CL, Miller MD et al. A comparison of ow-
33. Smith GE, Barnes R, Chetter IC. Randomized clinical trial of mediated dilatation and peripheral artery tonometry for
selective versus routine preoperative duplex ultrasound measurement of endothelial function in healthy individuals
imaging before arteriovenous stula surgery. Br J Surg 2014; and patients with peripheral arterial disease. Eur J Vasc Endo-
101: 469474 vasc Surg 2013; 45: 263269
34. Caroli A, Manini S, Antiga L et al. Validation of a patient-spe- 53. Hansell J, Henareh L, Agewall S et al. Non-invasive assess-
cic hemodynamic computational model for surgical plan- ment of endothelial functionrelation between vasodilatory
ning of vascular access in hemodialysis patients. Kidney Int responses in skin microcirculation and brachial artery. Clin
2013; 84: 12371245 Physiol Funct Imaging 2004; 24: 317322
35. Roy-Chaudhury P, Lee TC, Munda R. Predicting dialysis vascu- 54. Ortiz A, Massy ZA, Fliser D et al. Clinical usefulness of novel
lar access blood ow and diameter: too much, too little, or prognostic biomarkers in patients on hemodialysis. Nat Rev
just right. Kidney Int 2013; 84: 10761078 Nephrol 2011; 8: 141150
36. Menegazzo D, Laissy JP, Durrbach A et al. Hemodialysis access 55. Zoccali C. Asymmetric dimethylarginine (ADMA): a cardiovas-
stula creation: preoperative assessment with MR venog- cular and renal risk factor on the move. J Hypertens 2006; 24:
raphy and comparison with conventional venography. Radi- 611619
ology 1998; 209: 723728 56. Brister SJ, Buchanan MR. Effects of linoleic acid and/or marine
37. Grobner T. Gadoliniuma specic trigger for the develop- sh oil supplements on vessel wall thromboresistance in pa-
ment of nephrogenic brosing dermopathy and nephrogenic tients undergoing cardiac surgery. Adv Exp Med Biol 1997;
systemic brosis? Nephrol Dial Transpl 2006; 21: 11041108 423: 275278
38. Heye S, Fourneau I, Maleux G et al. Preoperative mapping for 57. Ridker PM, Rifai N, Rose L et al. Comparison of C-reactive
haemodialysis access surgery with CO(2) venography of the protein and low-density lipoprotein cholesterol levels in the
upper limb. Eur J Vasc Endovasc Surg 2010; 39: 340345 prediction of rst cardiovascular events. N Engl J Med 2002;
39. Hyland K, Cohen RM, Kwak A et al. Preoperative mapping 347: 15571565
venography in patients who require hemodialysis access: 58. Verma S, Wang CH, Li SH et al. A self-fullling prophecy:
imaging ndings and contribution to management. J Vasc C-reactive protein attenuates nitric oxide production and inhi-
Interv Radiol 2008; 19: 10271033 bits angiogenesis. Circulation 2002; 106: 913919
40. Patel ST, Hughes J, Mills JL Sr. Failure of arteriovenous stula 59. Verma S, Buchanan MR, Anderson TJ. Endothelial function
maturation: an unintended consequence of exceeding dialy- testing as a biomarker of vascular disease. Circulation 2003;
sis outcome quality Initiative guidelines for hemodialysis 108: 20542059
access. J Vasc Surg 2003; 38: 439445; discussion 445 60. Lee ES, Shen Q, Pitts RL et al. Vein tissue expression of matrix
41. Planken RN, Leiner T, Nijenhuis RJ et al. Contrast-enhanced metalloproteinase as biomarker for hemodialysis arterioven-
magnetic resonance angiography ndings prior to hemodi- ous stula maturation. Vasc Endovascular Surg 2010; 44:
alysis vascular access creation: a prospective analysis. J Vasc 674679
Access 2008; 9: 269277 61. Dunlay SM, Gerber Y, Weston S et al. Prognostic value of bio-
42. Bode AS, Planken RN, Merkx MA et al. Feasibility of non- markers in heart failure: application of novel methods in the
contrast-enhanced magnetic resonance angiography for community. Circ Heart Fail 2009; 2: 393400
imaging upper extremity vasculature prior to vascular access 62. Gerrits EG, Smit AJ, Bilo HJ. AGEs, autouorescence and renal
creation. Eur J Vasc Endovasc Surg 2012; 43: 8894 function. Nephrol Dial Transplant 2009; 24: 710713
43. Merkx MA, Huberts W, Bosboom EM et al. The benet of non 63. Lin CC, Chang CF, Lai MY et al. Far-infrared therapy: a novel
contrast-enhanced magnetic resonance angiography for pre- treatment to improve access blood ow and unassisted
dicting vascular access surgery outcome: a computer model patency of arteriovenous stula in hemodialysis patients. J
perspective. PLoS One 2013; 8: e53615 Am Soc Nephrol 2007; 18: 985992
44. Merkx MA, Bosboom EM, Bode AS et al. Non contrast-en- 64. De Marco Garcia LP, Davila-Santini LR, Feng Q et al. Primary
hanced MRA versus ultrasound blood vessel assessment to balloon angioplasty plus balloon angioplasty maturation to
determine the choice of hemodialysis vascular access. J Vasc upgrade small-caliber veins (<3 mm) for arteriovenous s-
Access 2013; 14: 348355 tulas. J Vasc Surg 2010; 52: 139144
45. Boutouyrie P, Fliser D, Goldsmith D et al. Assessment of arter- 65. Veroux P, Giaquinta A, Tallarita T et al. Primary balloon angio-
ial stiffness for clinical and epidemiological studies: metho- plasty of small (2 mm) cephalic veins improves primary
dological considerations for validation and entry into the patency of arteriovenous stulae and decreases reinterven-
European Renal and Cardiovascular Medicine registry. Nephrol tion rates. J Vasc Surg 2013; 57: 131136
Dial Transplant 2014; 29: 232239 66. Turmel-Rodrigues LA. Mechanical enhancement of AVF mat-
46. Vita JA, Hamburg NM. Does endothelial dysfunction contrib- uration. J Vasc Access 2014; 15(Suppl 7): S55S59
ute to the clinical status of patients with peripheral arterial 67. Lee T, Ullah A, Allon M et al. Decreased cumulative access
disease? Can J Cardiol 2010; 26: 45A50A survival in arteriovenous stulas requiring interventions to
47. Ganz P, Vita JA. Testing endothelial vasomotor function. Circu- promote maturation. Clin J Am Soc Nephrol 2011; 6: 575581
lation 2003; 108: 20492053 68. Dember LM, Imrey PB, Beck GJ et al. Hemodialysis Fistula
48. Thambyrajah J, Landray MJ, McGlynn FJ et al. Abnormalities Maturation Study Group. Objectives and design of the hemo-
of endothelial function in patients with predialysis renal dialysis stula maturation study. Am J Kidney Dis 2014; 63:
failure. Heart 2000; 83: 205209 104112
49. Owens CD, Wake N, Kim JM et al. Endothelial function predicts
positive arterial-venous stula remodeling in subjects with Received for publication: 2.1.15; Accepted in revised form: 5.3.15

Downloaded from https://academic.oup.com/ckj/article-abstract/8/3/282/405200/Current-tools-for-prediction-of-arteriovenous


by guest
on 21 October 2017