Patients with acute and chronic renal failure require dependable access for dialysis. Dialysis access failure has been reported to be one of the most frequent causes of hospitalization among patients with end-stage renal disease (ESRD).¹ With our ability to treat ESRD, improving the longevity of our patient population has been steadily increasing.

The Kidney Dialysis Outcomes Quality Initiative (DOQI),² as published by the National Kidney Foundation, sets forth recommendations as part of a national consensus that parishioners avoid percutaneous-catheter based arteriovenous (AV) hemodialysis access in favor of autogenous access (AA), followed by prosthetic access (PA), as a second preference. With vascular access (VA) complications accounting for 15% of hospital admissions among hemodialysis patients,^3,4 and Medicare costs approximating $182 million in 2003,⁴ the population of patients requiring hemodialysis access is expected to increase by 10% per year from a group which exceeded 345 000 patients in 2000.⁵

The current DOQI recommendations for practice patterns are the insertion of an AA in 50% of long-term access patients. However, some centers have had trouble achieving this goal as a result of vein mapping results or availability of forearm basilic vein.⁶ The DOQI guidelines-recommended surgical referral pattern should begin when a patient exhibits a creatinine clearance of less than 25 mL/min or a serum creatinine greater than 4 mg/dL or when AV access is anticipated within 1 year.²

The introduction of hemodialysis as routine treatment of ESRD made it necessary to find a simple form of repeated access to the vascular system. It was only after the introduction of external silastic cannulae by Quinton and Scribner⁷ in 1960 that extracorporeal treatment could be established. Several years later, Brescia and Cimino⁸ devised the AV fistula, which overcame the limitations of frequent infections and thrombosis. In the 1970s the implantation of grafts was introduced,^9,10,11 which permitted renal replacement therapy in patients devoid of venous vessels.

Currently, complications of VA (i.e., dysfunction, thrombosis, or infection) are a major cause of hospital admission. They affect the quality of life. For this there are objective reasons (they make it difficult to administer sufficient dose of dialysis) and subjective ones (anxiety because of uncertainty about correct functioning).¹² Furthermore, they give rise to frustration in health care personnel.^13,14,15 Recently, repeated VA failure has been identified as a risk factor for mortality.¹⁶ Finally, VA failure causes high economic costs, accounting for up to one-third of ESRD expenditure.¹⁷

The radiocephalic AV fistula is the preferred VA because of its low complication rates, its long survival, and its ease of puncture once it has matured.^18,19,20 Nevertheless, its establishment on the wrist or in the anatomical snuffbox of the nondominant arm is potentially inconvenient for two reasons: (1) Four to eight weeks are necessary until the venous wall has arterialized and (2) a high rate, 8% to 30% of initial failure or insufficient development is observed, necessitating the use of other modalities of VA.²¹ Recently, AV fistulae higher up in the forearm and on the upper arm have been put forward as acceptable alternatives. Some studies have documented primary patency rate >80% in the first 2 years of observation,²² but there is no information on the long-term outcome for this type of AV fistula.

The second mode of permanent VA are grafts, the use of which has increased in recent years, and in numerous centers it is today the most frequently used type of VA.²³ This tendency has been related to recent demographic changes in the hemodialysis population, the scarcity of transplants with the consecutive increased time on hemodialysis treatment, and increase comorbidity of patients beginning renal replacement therapy. In fact the median age of incident patients is actually around 60 years, more than half of the patients have at least two comorbid conditions, and 20% to 40% are diabetic patient—all factors that could affect the success of the VA.^17,24 On top of this, patients are not infrequently referred to the nephrologists in the terminal stages of renal failure or during an episode of acute deterioration of pre-existing renal failure. In these circumstances, it is frequently impossible to create a VA in time.²⁵

For patients who require immediate dialysis access, that is, those requiring hemodialysis of less than 3-week duration, a double-lumen cuffed or noncuffed catheter should be inserted into the femoral, internal jugular, or subclavian vein.²⁶ The most common catheter for this purpose is the Quinton catheter, which can be placed at the bedside and must be able to support a flow rate of 250 mL/min.²⁷ In case of femoral vein catheter insertion, the catheter should not remain longer than 5 days, because of the high propensity for infection or dislodgment with ambulation.²⁸ Most importantly, the subclavian position should be avoided if the patient is to be considered for an ipsilateral arm access procedure because the incidence of subclavian vein stenosis or thrombosis or both increases steadily with the presence of a catheter in this position, rendering the extremity useless for insertion of a permanent access.^29,30

For the patient who requires hemodialysis for more than 3-week duration, insertion of a cuffed, tunneled, double-lumen catheter should be considered. The preference of location would be the internal jugular vein. The right internal jugular vein is preferred because of its proximity to the atrial-caval junction (allowing for better flow), but with the added emphasis of placing the catheter in the right internal jugular.

Aside from complications associated with insertion, (hemothorax and pneumothorax), the tunneled cuffed catheter can be relied on to function for an average of 6 months, after which infection, fibrin sheath formation, or thrombosis may curtail usage.³¹ Using endoluminal therapy or percutaneous mechanical techniques, prolonged usage can be obtained up to 12.7 months as reported in a study.³² Local infection and sepsis,³¹ and infection elsewhere³³ are typical reasons for removal of the catheter.

Snuffbox Fistula

These consist of end-to-side anastomosis between the distal cephalic vein and the thenar branch of the radial artery, the pulse of which is usually palpable through the floor of the anatomic snuffbox, created through the one incision. In one European study over a 12-year period, 11% thrombosed within 24 hours of creation, and 80% had matured for hemodialysis within 6 weeks. The 1- and 5-year patency rates were 65% and 45%. Of the fistulae that thrombosed, ipsilateral wrist angio access was successfully constructed in 45%.³⁴ Similar results were obtained in another study,³⁵ which reported on 139 patients who underwent snuffbox fistula creation with and without diabetic nephropathy. After 57 months, 87% of patients without diabetic nephropathy had patent access, whereas 72% were patent among patients with diabetic nephropathy, and they concluded that patients with diabetic nephropathy may not arterialize their accesses as well as patients without diabetic nephropathy (Figure 48-1).

Radial-Cephalic

Direct autogenous radial artery-cephalic vein fistula was first described in 1966.³⁶ This access has also been called the Cimino fistula or the wrist fistula. Different configurations of anastomosis have been employed with varying results as far as development of a steal phenomenon or speed of maturation,³⁷ although the cephalic vein end-to-side configuration seems to be the most popular. Cephalic veins of less than 1.6 mm in diameter have been associated with early failure³⁸ (Figures 48-2 and 48-3).

Results of the Cimino fistula have been generally good, with 6-, 12-, and 36-month patency rate of 80%, 71%, and 64%, respectively.³⁹ In a series, reported factors suspected to have influence on AV fistula included age, BUN, blood pressure at the time of operation, serum cholesterol, and creatinine levels. The 1-year patency rate was 59.8% and 40.8% at 5 years.⁴⁰

Other reports suggest that early failure is considered one of the major determinants affecting the long-term patency, because, after 1 year, the rate of access failure was slow, steady, and near the same regardless of the type of surgery of the fistula.⁴¹

Our routine practice of creating a fistula technique is as follows:

1. 
   

   One incision at the wrist between the distal radial artery and cephalic vein under local anesthesia.

   
   
2. 
   

   Between the distal radial artery and the cephalic vein.

   
   
3. 
   

   Divide the distal end of the cephalic vein, after the patient is heparinized.

   
   
4. 
   

   Bring the end of the cephalic vein to the site of the radial artery in an end-to-side fashion with an arteriotomy of 5 to 8 mm in length.

   
   
5. 
   

   With the radial artery cross-clamped proximally after the arteriotomy is made, release the distal clamp of the radial artery. If pulsatile backflow is observed, we then ligate the radial artery distal to the end-to-side anastomosis. This pulsatile flow indicates complete palmar arcade.

By doing these techniques, we are able to improve on the following important points:

1. 
   

   Maturation is fast.

   
   
2. 
   

   Minimize early thrombosis because of flow-related problems.

   
   
3. 
   

   Minimize distal venous hypertension if you want to construct side-to-side anastomosis.

   
   
4. 
   

   Minimize distal ischemic change because of reversal of the flow from the distal radial artery into the vein at the later stage when the venous system becomes enlarged.

Anastomosis of the antecubital veins with the brachial artery can be accomplished with good results. Despite favorable results, the fistula has a higher incidence of steal, especially with long donor arteriotomies. In one study, the antecubital vein fistula had a primary patency rate of 80% at a medium follow-up of 36 months, compared with 66% of brachial–cephalic fistulae at 24 months.⁴² This study suggests that the brachial–cephalic fistula was a favorable alternative in elderly patients, women, and diabetic patients. In another study, 74% 1-year patency rate was accomplished.⁴³

Brachial–Radial at Forearm

Hakaim et al.⁴⁴ showed superior patency and maturation rates of primary brachial–basilic vein with transposition, compared with 78% and 79% for brachial–basilic and transposed brachial–basilic access.

Brachial–Basilic at Upper Arm

This procedure was first described in 1976.⁴⁵ The procedure involved mobilization, distal division, and superficial tunneling and transposition of the basilic vein with distal end-to-side anastomosis with the brachial artery. The technical alternatives and modification include elevating the basilic vein rather than rerouting it⁴⁶ or superficializing the basilic vein as part of a staged procedure.⁴⁷ Others employed endoscopic techniques as a means of reducing the incision length.⁴⁸

In a large study, the long-term follow-up for patients having undergone the procedure were, 1-year patency of 84%, 73% at 3 and 5 years, and 52% at 10 years.⁴⁹ Also, less favorable results were reported in other studies.⁵⁰

In a rare incidence, lower extremity AA can be done. This technique involved dissection and mobilization of the entire superficial femoral vein, with transposition into a superficial position in the thigh, and end-to-side anastomosis.⁵¹ In a retrospective analysis of 25 patients for more than 2 years, cumulative patency was 78% and 73% for 6 and 12 months follow-up. Steal syndrome necessitated further intervention in 40%, and of those, 80% required another procedure to treat steal. Major wound complications affected 28%, and one patient required above-knee amputation after developing a compartment syndrome. Therefore, it has been suggested not to undertake this technique without serious evaluation, such as good-risk patients who have no other possible sites for fistula creation.

There are currently 2 major manufacturers of expanded polytetrafluoroethylene (PTFE) for use in hemodialysis access; Gore-Tex (W. L. Gore & Associates, Flagstaff, AZ) and Impra (Impra, Inc., Tempe, AZ). Although manufacturer claim that each of these are having distinct patency and cost advantages over the other, this has not been borne out comparative investigations of either product.^52,53

Stenosis of the venous outflow, generally as a result of neointimal hyperplasia, remains the sentinel cause of graft failure, accounting for approximately 80% of graft failure.⁵⁴ Measurement of the outflow tract have been correlated with graft patency: lesions that account for less than 30% stenosis were associated with a less than 30% thrombosis rate at 6 months, whereas lesions that accounted for greater than 50% of the outflow were associated with an almost 100% failure rate at 6 months.⁵⁵ As a result of the association of turbulent flow at the anastomosis with the formation of neointimal hyperplasia,^56,57 investigators designed expanded PTFE grafts that incorporated a cuffed geometry of the venous anastomosis site, a design that had shown bench utility in minimizing shear stress.⁵⁸ This configuration was studied prospectively in 48 patients, although overall primary patency was not affected, secondary patency was increased from 32% to 64% at 12 months.⁵⁹

There are many varieties of prosthetic angio access graft procedures.

Forearm, Loop (U-Shaped)

Technique

A small transverse incision is made on the upper portion of the forearm, approximately 2 fingerbreadths distal to the antecubital vein. The superficial vein and antecubital vein are isolated. Depending upon the condition of the antecubital vein, sometimes one has to use the basilic portion or a separate portion proximally, or at the antecubital vein. Using the same incision, the brachial artery is isolated. Then a subcutaneous tunnel is made distally to the distal forearm near the wrist curving back to the antecubital incision in a U-shape. Small counted incisions at the distal forearm are made to facilitate the tunnelling procedure with the ease of bringing the graft into the tunnel. After appropriate heparinization, we normally make an end-to-side anastomosis on the venous side first, using either 6.0 or 7.0 PTFE sutures. The graft is then passed through the tunnel into the antecubital incision. At this point, an arterial anastomosis is made in an end-to-side fashion. Following the completion of the procedure, we either neutralize the heparin or leave the heparin on board with careful hemostasis (Figure 48-4).

Forearm, Straight Graft

Using the brachial artery or the radial artery and the antecubital vein or distal cephalic vein, the graft is placed in the subcutaneous tunnel.

Upper Arm, Loop (C-Shaped)