Criteria for Accessory Vein Obliteration Trial.

Hemodialysis (HD) vascular access dysfunction is a huge clinical problem which results in significant morbidity amongst patients with End Stage Renal Disease (ESRD), causing a severe economic burden on any health care system. Native arteriovenous fistula (AVF) is the preferred form of permanent dialysis access since it is associated with the best long term outcome as compared to other forms of dialysis access. Despite the clinical benefits, many patients on HD do not have AVF as there dialysis access. One of the reasons for low AVF rates is early fistula failure (EFF). EFF is defined as an AVF that never develops adequately for dialysis (failure to mature) or which fails within 3 months of starting dialysis. An adequate AVF for dialysis according to Dialysis Outcome Quality Initiative (DOQI) guidelines is the one which a) Has a flow of greater than 600ml/min, b) Has a diameter of 0.6cm or greater and c) Is approximately not deeper than 0.6cm from the skin surface. Between 23%-46% of newly constructed AVF have problems with early failure resulting in a dismal one year patency of 60-65%.

In order to devise a strategy to prevent EFF, one needs to understand the physiology of fistula maturation. Creation of an AVF leads to an immediate increase in flow through the vein due to the pressure gradient created. This increase in flow leads to increase wall shear stress which is defined mathematically by the formula 4ηQ/πr3, where η is blood viscosity, Q is blood flow and r is vessel radius. Shear stress thus is directly proportional to blood flow while inversely proportional to vessel diameter. After the creation of the AVF, the flow mediated increase in shear stress is mitigated by vessel dilatation through biological mediators. Consequently the shear stress is brought back to pre-anastomosis levels leading to vessel dilation. It seems that this positive remodeling of the vein leading to AVF maturation is dependent on increase in blood flow rather than the increase in pressure. Any pathology affecting the blood flow through the newly constructed AVF can thus lead to failure to mature.

The two most important causes for EFF amenable to intervention are stenosis anywhere in the circuit and/or presence of accessory veins. While stenosis development is pathological and accessory vein (av) presence is natural, both lead to decreased blood flow through the main AVF circuit, which may be responsible for failure to mature. Addressing these two entities in a timely fashion can lead to salvage of many AVF, which otherwise would have been abandoned. Although management of stenosis is well established with relatively clear guidelines, the management of av lacks clear scientific approach. It has been anecdotally recommended that any av with a diameter greater than 25% of the AVF diameter should be considered for ligation. It is also suggested that the significance of an av can be ascertained by assessing the flow of contrast through the av during an angiogram or by assessing the augmentation of AVF after manually occluding the av during the procedure. These current recommendations are very subjective and none have been adequately studied to date leading to considerable debate on the issue of when to obliterate an av. A scientific approach to obliteration of av is essential to avoid an unnecessary procedure, which may even be harmful. Also, obliterating a non-significant av and waiting for AVF to mature may cause a delay in making a definite decision for the patient. These patients may have other factors and not av as the cause of EFF. Conversely, not obliterating a significant av may also delay the use of AVF for dialysis by causing EFF. In a recent study researchers recommended a hemodialysis arteriovenous flow quantification-diameter (HAQ) criteria for accessory vein obliteration written below:

1. 60% or greater diameter of the main AVF
2. 50% diameter of AVF with at least one more av>40% in diameter.
3. 50% in diameter and divides into branches of same size.
4. av likely to interfere with cannulation on physical examination.
5. >30% in diameter and associated with stenosis at site of origin.

The HAQ criteria for av obliteration is based on computational flow dynamic (CFD) models of AVF created using 2 dimensional CFD software. This criteria has not been tested in clinical trials and its clinical utility has not been established to date. The purpose of this controlled, blinded, prospective trial is to assess the clinical utility of the HAQ criteria as compared to current recommendations for av obliteration.

Specific Aim:

The purpose of this controlled, blinded, prospective trial is to assess the clinical utility of the HAQ criteria as compared to current recommendations for av obliteration.

Method and Materials:

Trial Design: The study will be a single center blinded, controlled and prospective trial. The study will be conducted according to the ethical standards of King Faisal Specialist Hospital \& Research Center (KFSHRC) for human experimentation and in compliance with the declaration of Helsinki.

Participants: The study will be conducted at the Interventional Nephrology section at KFSHRC, Riyadh, Saudi Arabia. All adult patients (Age>18 years) with End Stage Renal Disease (ESRD) or chronic kidney disease (CKD) with estimated glomerular filtration rate less than 10 approaching ESRD who have had an AVF created and has not matured enough to be used for dialysis 6 weeks after surgery will be eligible for consideration.

Design: All participants with EFF are currently evaluated for cause of EFF as the standard of practice at KFSHRC. Patients who agree to participate by providing informed consent will undergo the same routine evaluation for EFF. As mentioned before, stenosis anywhere in the AVF circuit and presence of av are the 2 most common causes for EFF. Patients in whom stenosis is the sole cause of EFF will be excluded from the study. Amongst eligible patients in whom av is the only cause of EFF will be directly undergo av obliteration. Before and after any av obliteration assessment of av will be done to ascertain whether it satisfies current recommendation or the HAQ criteria and to assess the effect of obliteration on access flows as shown in data collection sheet. Data on av will be interpreted by another expert, blinded to study details (other than the operator) after the procedure to assure accuracy. Those patients in which both av and stenosis are present will first undergo angioplasty of the stenotic region. These patients will be brought back at 4 weeks post intervention for re-assessment of AVF maturation by staff blinded to intervention details. av will be managed if indicated at that time. Patients who undergo av obliteration will be reassessed in 4 weeks for AVF maturation by staff blinded to intervention details. Patients who are thought to have mature AVF on clinical exam will be scheduled for cannulation. Final maturity will be defined as an AVF which undergoes 3 consecutive successful HD treatments using the AVF. Access flows will also be measured. The access flow studies using ultrasound dilution technique can be done on either of the first three treatment sessions. Ultrasound dilution technique is an accepted method for assessing access blood flows. Patients with mature AVF following interventions will be followed for 6 months with monthly evaluation with access flows and physical examination to assess primary and secondary patency. The follow up can be extended to another 18 months to assess long term patency but during that time access management and surveillance will be as per local center practice.

Procedural Details:

Angiograms under fluoroscopy will be performed as standard practice in such procedures. To assess arteriovenous anastamosis stenosis, arteriograms will be done rather than reflux angiogram which is more accurate in assessing av. As part of current guidelines for practice a puff of contrast is injected in the AVF and if there is av present it is assessed subjectively without any measurement and if it is thought to be significant it is ligated. But in this study the investigators will take measurements of the av as well as of the AVF which is the only additional step from the standard of care .In case of av presence, downstream sheath placement from the origin of av will be avoided as this may cause functional downstream stenosis leading to slightly high pressure in AVF causing alteration in the size and flow through the av. Data on the size of av and AVF diameter will be collected after adequate calibration of the system. This will be collected and sealed before any intervention and later interpreted by another expert in the field blinded to intervention to assure accuracy. This interpreter will assign whether the av satisfies the Standard or HAQ criteria.The investigators currently use a Phillips veradius fluoroscopy equipment. If the av qualifies for obliteration than obliteration will be performed. For obliteration of av either the surgical ligation or coil embolization will be used. Both are established current standards of practice. Successful obliteration will be defined as complete disappearance of av on repeat angiogram following the procedure. For details of data collected please see the data collection sheet at the end.

Statistical Design:

Sample Size:

For the purposes of designing this study in terms of sample size, the outcome variable of interest is whether the fistula matures or not, i.e. a binary outcome. The associated parameter is the so-called binomial parameter and is denoted here as π. The predictor variable is whether the patient meets the HAQ criteria or not, i.e. also a binary variable. The notation for this variable will be X, and that will take on the value one if the patient meets the HAQ criteria and the value zero if the patient does not meet the HAQ criteria. Given this notation, then π will be further defined as π0 for those patients that do not meet the criteria, and π1 for those who do.

In statistical terms, the research hypothesis is H0: π1 - π0 = 0 vs H1: π1 - π0>= δ> 0. Of clinical interest is to conclude H1 if δ is at least equal to 0.20, i.e. that the probability that the fistula will mature is at least 20% greater for those meeting the HAQ criteria than for those not meeting it.

With the research hypothesis framed as such, then the sample size calculation becomes one of estimating a binomial parameter, i.e.

n = 4(1-πi)πi/M squared, where here M equals 0.05. For varying values of πi, n is maximized for πi equal to 0.5. At such a maximum, then n equals 25, i.e. 25 patients would be required of each type (those meeting and those not meeting the HAQ criteria).

Statistical Analysis According to the above description of the statistical design of the project, the analyses will involve a test between two binomial proportions. Fisher's Exact test would be carried out and with a size α = 0.05. Additional analyses will also be carried out to investigate the exact relationship between the size of the accessory vein and the probability of the fistula maturing. This would be a logistic regression analysis with repeated covariates and will be analyzed with generalized estimating equations (GEE).