Vascular access is required to assist in the man- agement of patients requiring frequent venous or arterial cannulation. Its principal role is to provide circulatory access for hemodialysis, al- though such techniques are also used to pro- vide chemotherapy, intravenous nutrition, and access for plasma exchange.
The two techniques used are either implant- able synthetic lines, which may have one or more lumina, or a surgically created fistula between the arterial and venous circulation.
This latter approach may involve joining artery and vein together, the so-called autogenous arteriovenous (AV) access [arteriovenous fistula (AVF)], or the insertion of a synthetic bridge graft between artery and vein (nonautogenous AV access).
Principles of Vascular Access
The guiding principles of access surgery are to use AVF in preference to synthetic grafts, which are preferred to in-dwelling central venous catheters. Access should be sited as far distally in the chosen limb as possible. The principle is to conserve veins to permit proximal revision if the initial procedure fails. An exception is very old patients, who often have poor distal vessels.
Some young patients may have strong wishes to avoid visible forearm scars. The nondominant limb should be used wherever possible.
In patients who may require vascular access, venepuncture, or the insertion of intravenous
cannulae in the main named veins of the upper limb, should be avoided. If central venous catheters are required, they should be sited in the jugular and not the subclavian veins to avoid central venous stenosis in the draining veins of the upper limb.
Standards
The United States has led the way in this area with the publication of specific guidelines [National Kidney Foundation-Disease Out- comes Quality Initiative (NKF-DOQI) (www.
kidney.org/professionals/doqi/guidelines)].
These guidelines were a response to the rapid growth in patient numbers and costs associated with treatment for renal failure. A multidiscipli- nary group of nephrologists, nurses, vascular and transplant surgeons, and interventional radiologists reviewed the world literature. The resulting document was approved by all the rep- resented national societies prior to adoption.
The principal recommendations are that the majority of patients starting dialysis should do so using a native vessel AVF at the wrist. Grafts require six times as many interventions as AVF to achieve the same patency, and their use is dis- couraged.Central catheters have inferior patency and are discouraged for “permanent” access.
Their principal role should be for emergency or temporary access. Since the distribution of these standards,placement rates for simple fistulas has increased 35% in the United States.
Vascular Access
David C. Mitchell and C. Keith Ozaki
141
The Europeans have not produced a similar document, although it is hoped that a National Service Framework for renal disease manage- ment in the United Kingdom will set standards for vascular access within the National Health Service.
Observational studies have helped to define standards. The Dialysis Outcomes and Practice Patterns Study (DOPPS) was a prospective global study of practice in seven countries involving 309 facilities and over 9500 patients.
It showed that there are significant differences between varying health cultures in the devel- oped world. The AVF rates varied from 0% to 87% in the United States, and 39% to 100%
in Europe. Graft usage in new hemodialysis patients ranged from 2 to 24%. Central venous lines were particularly common in the United Kingdom and the United States but rare in Japan and Italy.
These differences are not dictated solely by differences between patients, but principally by clinician preference. This study has clarified what is being done at present to provide access.
By demonstrating widely differing practice, it informs the debate over ideal practice and can guide future trials to answer questions about treatment where widely differing views are held.
Interpretation of the literature is further com- plicated by differences in reporting of the re- sults of access surgery. A recent study from the United States should help to standardize report- ing in the future (Sidawy et al., 2002).
Types of Access
The principal modes of vascular access are autogenous AVF, synthetic AV bridge grafts (nonautogenous AV access), and in-dwelling synthetic central venous catheters. Considera- tion of patient demographics, anatomical suit- ability, and other factors are important in the preoperative planning for access (Table 13.1).
Arteriovenous Fistula
Most clinicians agree that the most durable form of vascular access is the AVF. The prefer- ential site is between the cephalic vein and the radial artery at the anatomical snuffbox, or the wrist, in the nondominant upper limb, the
Brescia-Cimino fistula (Brescia et al., 1966). The end of the vein is joined to the side of the radial artery, usually under local anesthesia.
The success of this procedure is dependent on the quality of the vein and artery and the tech- nical skill of the surgeon. Where good vessels are not clinically evident, duplex ultrasonogra- phy or venography may help to identify the best sites for access formation. The suggested sequence for AVF placement is listed in Table 13.2).
Widely varying success rates are quoted, but about 60% of wrist AVFs mature and become useful for dialysis. The AVFs are robust once established, but may take many (typically 6 to 12) weeks to mature. Many factors may affect Table 13.1. Preoperative planning for autogenous or nonautogenous arteriovenous fistula (AVF) placement
Consider patient age, projected life span, time to initiation of dialysis, dominant hand
Elicit history of previous access attempts/subclavian vein cannulations; consider contrast venography to rule out central venous pathologies
Check Allen’s test, upper extremity blood pressures, arterial pulse exam Æ if abnormal then segmental arm pressures, plethysmography Æ if abnormal then consider ultrasound or angiography Venous exam with tourniquet Æ if no clear conduit
then duplex; protect that vein from puncture/trauma
Ensure patient is optimized for operation (e.g., address cardiac, metabolic, volume status, nutritional, infectious issues)
Consider risks of steal in elderly and diabetic patients undergoing proximal access construction
Table 13.2. Suggested sequence for AVF placement 1. Autogenous AVF in hand/forearm (nondominant
before dominant)
2. Upper arm autogenous AVF (usually cephalic before basilic)
3. Forearm nonautogenous AVF 4. Upper arm nonautogenous AVF
5. Upper arm or thigh autogenous (using transposed or translocated saphenous/superficial femoral vein) AVF
6. Thigh nonautogenous AVF
7. Central configurations such as axillary artery to contralateral axillary vein, subclavian artery to subclavian vein
maturation, such as the size of artery and vein, change in flow, and the presence or absence of arterial disease (e.g., diabetes). In patients in whom the fistula is failing to become prominent within a few weeks, duplex ultrasound scanning can examine flow rates and identify problems.
Surgeons willing to undertake revision may produce higher success rates. The principal reasons for failure are damaged or inadequate veins and (more rarely) inadequate arteries. In such cases, alternative forearm veins may be used. The basilic vein on the ulnar side of the forearm is often large and may not have been traumatized by venesection. This can be mobi- lized and swung across the forearm to the radial artery, or anastomosed to the adjacent ulnar artery if large enough, although the failure rates for this procedure are higher.
Any vein can be damaged by repeat venesec- tion or placement of in-dwelling catheters. Such veins develop areas of fibrosis that cannot dilate when subjected to arterial flows. For this reason, in those in whom the need for access can be anti- cipated, every effort should be made to avoid needling the cephalic vein, the antecubital veins, or the subclavian vein.
The radial artery at the wrist is often insuffi- cient in the elderly and particularly those with diabetes. The artery usually shows evidence of arteriosclerosis with calcification. The artery is unable to increase its flow rate in response to fistula formation. The result is failure of the vein to enlarge, or thrombosis. In this situation, it may be better to place the fistula more proxi- mally in the limb, typically at the elbow. Pre- operative evaluation can assist in identifying suitable patients for distal AVFs (Malovrh, 2002).
The standard procedure at the elbow is an end of cephalic vein to side of brachial artery fistula. This can usually be fashioned through a transverse antecubital incision or two lon- gitudinal incisions under local anesthesia. The depth of the basilic vein within the arm com- plicates its use as a direct fistula. This vein, however, may be easily mobilized through a lon- gitudinal incision in the medial arm. The basilic vein is then divided distally, tunneled subcuta- neously, and anastomosed to the brachial artery in the distal arm. The procedure can be per- formed using minimally invasive techniques.
Such transposed fistulas may give good service over many years. This technique may also be
applied to the lower limb in patients with good arterial circulation, by transposing the long saphenous vein subcutaneously in the thigh and anastomosing it to the distal superficial femoral artery.
Synthetic Access Grafts (Nonautogenous Arteriovenous Access)
Where veins are inadequate, or dialysis is needed urgently (within 2 weeks), AVF may be inappropriate. In such circumstances, access can be rapidly established using synthetic bridge grafts between suitable arteries and veins. The most common material for this is polytetra- fluoroethylene (PTFE), which may be cannu- lated within a week if required. Dacron is not widely used for access as it is difficult to needle, but newer composite grafts that are reported to bleed less after cannulation are available.
The role of these grafts has yet to be clearly established.
As veins are often deficient, grafts tend to be placed more proximally in the limb than AVFs. They may be placed in either straight or curved/looped configurations. No data exist demonstrating that one configuration is signi- ficantly better than any other. Grafts should be appropriately sized with the arterial anastomo- sis not bigger than 6 mm to avoid excessive flows and either vascular steal or high output cardiac failure.
Although easier to establish, PTFE grafts are known to have higher thrombotic and septic complication rates, and to need more frequent revision than native AVFs. For this reason, it is advised that AVFs are preferred. This has never been formally tested in a prospective random- ized trial. In the elderly patient with limited life expectancy, a rapidly established graft may provide superior access with sufficient durabil- ity, compared to several unsuccessful attempts to establish an AVF.
Central Venous Catheters
A well-organized renal failure service should anticipate the need for access in patients ap- proaching end-stage disease, and appropriate access surgery should be planned in advance.
This counsel of perfection is not always realized as patients may present acutely after a long period of stability.
Those requiring emergency access (e.g., im- mediate need for dialysis) need the insertion of a dual-lumen central venous catheter. These should be left in place for as short a time as pos- sible. Definitive access surgery should follow as a matter of urgency.
Central catheters become infected easily, and often have to be removed or re-sited. In addi- tion, they stimulate fibrosis in the veins and can cause stenosis. For this reason, subclavian lines are to be avoided, as the loss of the subclavian vein prejudices further access in the ipsilateral upper limb. Lines should be placed via the jugular route if at all possible. If access is needed only for a few days, then in the absence of suit- able jugular veins the femoral veins may be used.
Patients needing more permanent catheters (longer than 3 weeks) should have tunneled catheters inserted. Placement considerations are the same as for temporary lines. Tunneled catheters should be cuffed, as the cuffs reduce the risk of infection from the site of skin entry.
Complications of Vascular Access
Arteriovenous Fistula
Once the AVF is established, AVF revision rates are low at about 15% or less per annum.
Fistulas are robust and withstand multiple cannulations well.
Technical Failure
The most immediate complication of AV access is that the fistula occludes shortly after forma- tion. This may be due to obvious problems such as inadequate vessels or to technical imperfec- tions in the procedure. In the former, re-siting the fistula is the best course of action. In the latter situation, revision, often under local anesthesia, may salvage a functioning AVF. This is rarely an emergency and can be managed during the next available elective operating session.
Failure to Mature
Failure to mature is the most common problem seen in AVFs. Fistulas demonstrate significant increase in flow within 48 hours and enlarge thereafter. Most successful AVFs are capable of being needled in about 6 to 8 weeks after formation, although longer may be required.
Detecting those that will fail to mature may be difficult. Ideally, AVFs should be placed well in advance of the time that they are likely to be required. This permits time for assessment and revision if the fistula is not developing satisfactorily. An alternative approach if time is pressing is to scan the fistula with ultra- sound soon after formation. Flow rates greater than 500 mL/min signify that the fistula is likely to develop successfully. Scanning can also be used for the early detection of technical deficiencies, stenoses, and other problems. In constructions that are failing to mature, some groups advocate early liberal use of fistulograms to aid in diagnosis and therapeutic planning.
Others rely primarily on the diagnostic vascular laboratory, and reserve fistulography for cases in which duplex scanning is equivocal or unavailable.
Vascular Stenosis
Vein stenosis and aneurysmal dilatation are both seen following development of AVFs. The former may compromise both the quality of dialysis and the longevity of the fistula. Some centers advocate routine fistula scanning, but this has not yet been shown to be an effective way of monitoring fistulas for complications.
Where obvious stenoses exist, or dialysis is inadequate despite an apparently satisfactory fistula, then ultrasound may reveal a stenosis.
Stenoses greater than 50% or AVFs with flows below 500 mL/min are more likely to fail and may need revision. Stenoses within 1 to 2 cm of the anastomosis are most easily dealt with by surgical revision. This can be undertaken under local anesthesia with an occlusive tourniquet.
The fistula is ligated and divided and reanasto- mosed to an adjacent portion of artery. More remote stenoses may often be dealt with by angioplasty. If this fails, then short skip grafts of vein may bypass a stenosis and permit continu- ous use of the fistula without the need for tem- porary lines.
Vascular Steal and High-Output Cardiac Failure
Proximal fistulas tend to have higher flow rates and may divert blood from the hand. The risk of vascular steal should be borne in mind in proximal fistula. It is uncommonly seen (about 1% to 8%), but the neurological effects are irre- versible if the steal is not quickly corrected (Tordoir et al., 2004). Any suspicion of hand ischemia demands immediate attention, begin- ning with bedside examination. Temporary finger occlusion of the AVF can assist in deter- mining whether there is an inflow problem, or whether the arterial flow distal to the arterial anastomosis has been interrupted. If hand arte- rial perfusion returns to baseline with AVF com- pression, then inflow is inadequate to maintain flow to the hand and through the AVF. If the extremity remains ischemic, then there may be a problem with the arterial anastomosis that occludes outflow, or thromboembolic complica- tions. Segmental arm and finger pressures, fol- lowed by repeated exam while the access is occluded, confirm the diagnosis. If inflow is inadequate, then improvement of distal pres- sures and symptoms should occur while the AVF or graft is compressed.
If steal physiology is demonstrated, then appropriate urgent imaging to screen for occult inflow arterial occlusive disease may be under- taken, and lesions addressed. Fistula revision is often required. If preservation of the fistula is not essential, taking it down and creating a new one is probably the procedure of choice.
The alternatives are banding to reduce flow, or bypass from the artery at least 6 cm above the anastomosis to an artery distally and then liga- tion of the artery immediately beyond the anas- tomosis. This last procedure is known as the distal revascularization-interval ligation (DRIL) procedure. It is gaining in popularity as it takes flow to the distal arterial tree from above the fistula and prevents steal due to reversed flow in the artery distal to the anastomosis.
Banding is a complex procedure, necessitat- ing intraoperative flow measurement. Flows may fluctuate during the procedure, but ideally the banding should produce a flow between 800 mL and 1.5 L per minute. Success rates for this procedure are low, and it is not recommended if an alternative is available. Postoperative throm- bosis is common and may be due to overvigor-
ous banding. If this happens, the fistula is usually lost.
High fistula flows can also cause high-output cardiac failure, which is seen more commonly in proximal AVFs. In these cases, fistula revision to reduce flows is required. The choices include sacrificing the fistula and siting new access, or attempting a flow reduction procedure.
Infection
Infection is rare, but if associated with bac- teremia, abscess, or aneurysm formation, revi- sion is usually required. The clinical features are of an area of inflammation in relation to the fistula, sometimes with rapid enlargement.
Fever and rigors, particularly on dialysis, may be seen, but are not a universal feature. Sudden rapid bleeding can occur at infected needling sites. Immediate pressure and placement of a skin suture will achieve acute control, but should be followed by a definitive revision in most cases.
Thrombosis
Thrombosis is uncommon, but when it occurs it can often be retrieved by acute intervention, with salvage of the fistula and avoidance of tem- porary central lines. Overdialysis is an uncom- mon cause of fistula thrombosis, but should be suspected if no stenosis can be found. Imaging of the whole fistula and ipsilateral central veins should always follow surgical or radiological declotting. Dilatation of fistula or central vein stenosis may be required to restore normal function. When the stenosis is in close proxim- ity to the anastomosis, then surgical revision to an adjacent proximal arterial segment is often the best procedure.
The complications associated with trans- posed fistulas are the same as those seen with other native AVFs. They tend to have high flows, so cardiac failure and steal are more common in these fistulas.
Arteriovenous Access Grafts (Nonautogenous Arteriovenous Access)
Arteriovenous grafts can usually be established easily, with a high technical success rate. How-
ever, complications are common, and the fre- quent need for intervention should prompt the surgeon and dialysis team to undertake regular surveillance.
Thrombosis
Thrombosis is usually the consequence of intimal hyperplasia at the venous end of the graft. This can be recognized by regular intra- dialysis monitoring of flow or pressure. It has been shown that monitoring of flow or pressure can detect developing graft stenosis. Timely intervention can prolong graft survival. One approach is to use balloon angioplasty where possible, keeping surgical revision to a minimum. Where thrombosis occurs, prompt declotting undertaken by either an endovascu- lar or open approach can retrieve graft function.
It is important to identify and correct the cause of the thrombosis or further thrombosis is likely. There is no clear evidence whether open or endovascular approaches are superior in this situation, and each approach has its proponents.
As grafts require relatively frequent interven- tions, many advocate using minimally invasive techniques first and reserving open surgery for those in whom the vascular interventionist fails.
Infection
Like any foreign material, infection is an ever-present hazard, especially as needles are frequently introduced into the graft. Septic episodes in dialysis patients are most likely to originate from the graft. Great care must be taken to use aseptic techniques when cannulat- ing grafts, as once organisms are seeded into a graft the only way to get rid of them is to replace the graft. Eradication of nasal car- riage of Staphylococcus aureus by use of topical mupirocin has been shown to reduce the rate of graft sepsis.
Grafts are often placed in European prac- tice as a dialysis access of last resort, and there may be limited alternative sites if an infected graft is removed. Clearly if there is inflamma- tion along the length of the graft, then removal is the only option. Where localized sepsis exists, or a small area of the graft is exposed, local
excision with skip grafting through uninvolved tissue may keep the graft functioning. The site of infection should be left open to drain freely, and the patient treated with parenteral anti- biotics until the inflammation has completely subsided.
Vascular Steal
This is most commonly seen in arm grafts in older or diabetic patients. Rates vary from between 1% to 19%. The approach is the same as steal in an autogenous AVF, with nonautoge- nous graft ligation or DRIL being the potential management approaches.
Aneurysm Formation
With repeated cannulation over small areas of a graft, there is a tendency for “aneurysm” for- mation (actually a pseudoaneurysm). This is a result of longitudinal splitting of the graft. It is not a problem if there is good skin coverage, but if enlarging rapidly or if inflamed, then early revision should be undertaken to prevent rupture and bleeding. The technique is similar to that used for infection, with the exception that excision of the aneurysmal portion of the graft is not always necessary in the absence of infection.
Central Venous Catheters
Infection
Infection is the main problem from central venous (CV) catheters, and rigorous aseptic technique is required to manage dialysis. Newer, totally implantable catheters that are accessed through the skin may reduce the risk of sepsis.
The teaching of correct hand and skin prepara- tion prior to needling remains an important part of the management of patients with CV catheters.
Once infected, lines need to be changed.
Ideally, the old line should be removed and the patient treated until he or she is well before inserting a new line. This is not always possible.
Rotating the line to a new site is to be preferred in such situations, but if the situation is desper- ate, removing the old line and reinserting in the
same location while treating with antibiotics may be successful.
Fibrinous Encasement and Clotting
The tips of catheters become covered with a
“biofilm” of fibrin that may contain bacteria.
Even if sterile, this film can build up until a large tube of fibrin encases the catheter tip. This can interfere with blood flow and obstruct dialysis.
If the line is not infected, removal and replace- ment may still be necessary to reestablish dial- ysis. Passing guidewires down the catheter is not usually successful, but snaring the catheter tip radiologically and stripping the fibrin cuff can clear the offending plug. This technique can significantly prolong the life of central lines.
Thrombosis is a common problem, especially if the line is not used frequently. Lines should be flushed with the correct volume of heparin (1000 U/mL) after each use to remove blood from the lumen and prevent thrombosis. Extra care must be paid to this in children, as it is easy to give excessive volumes of heparin and induce systemic anticoagulation and bleeding.
The Avoidance of Complications:
Access Surveillance
Concerns about the frequent need for access revision have stimulated inquiry into tech- niques for the early detection of access com- plications before symptoms develop. Most techniques concentrate on intradialytic moni- toring of flow, pressure, dialysis efficacy, or a combination of these. The advantage of intradi- alytic monitoring is that it minimizes the need for extra hospital resources for these ill patients.
Each technique has its proponents, and the tech- niques mentioned are not mutually exclusive.
The DOQI guidelines state, “Access flow measured by ultrasound dilution, conductance dilution, thermal dilution, Doppler or other technique should be performed monthly. The assessment of flow should be performed during the first 1.5 hours of the treatment to eliminate error caused by decreases in cardiac output related to ultrafiltration. The mean value of
three separate determinations performed at a single treatment should be considered the access flow. If access flow is less than 600 mL/
min, the patient should be referred for fistulo- gram. Access flow less than 1000 mL/min that has decreased by more than 25% over 4 months should be referred for fistulogram.”
This is a counsel of perfection and some centers would advocate duplex ultrasound prior to fistulography. Where there is evidence of a stenosis of >50%, there is a significant sub- sequent thrombosis rate. Such fistula should undergo endovascular or surgical revision as appropriate. Juxta-anastomotic stenoses are often best dealt with by surgery, with endovas- cular dilatation being reserved for those in the body of the access.
Desperate Access
Dialysis is becoming more successful in pro- longing the lives of patients in renal failure.
Unfortunately, this improvement in survival is not being matched by increases in transplanta- tion. As a result, many more people are depend- ent on long-term dialysis. As successive access procedures fail, the establishment of secure vas- cular access becomes increasingly difficult.
Surgeons and interventionists have become ingenious at inserting grafts or lines into various veins, sometimes accidentally. Grafts can be placed in necklace fashion between one axilla and the contralateral one. Arteries can be divided and interposition grafts inserted.
The superficial femoral vein has been utilized as a hemodialysis access conduit. Lines may be inserted directly into the inferior vena cava (IVC) through the back. Such procedures should be reserved for those in whom no alternative can be found after an extensive search using ultrasound, venography, and magnetic reso- nance angiography if necessary. It should be made clear to patients undergoing these proce- dures that the risks involved in establishing such access are high, and the consequence of failure may be fatal.
Although successful arterial access can be obtained for numerous patients with reason- able durability, numerous controversies remain (Table 13.3). These controversies will be resolved with further clinical trials.
References
Brescia MJ, Cimino JE, Appell K, Hurwich BJ, Scribner BH.
(1966) J Am Soc Nephrol 10:193–9.
Malovrh M. (2002) Am J Kidney Dis 39:1218–25.
Sidawy AN, Gray R, Besarab A, et al. (2002) J Vasc Surg 35:603–10.
Tordoir JH, Dammers R, van der Sande FM. (2004) Eur J Vasc Endovasc Surg 27:1–5.
Table 13.3. Contemporary controversies
Role of endovascular vs. open management of the failing and failed autogenous AVF and nonautogenous AVF
Construction of upper arm autogenous AVF prior to utilization of forearm nonautogenous AVF Role of invasive imaging (routine arteriograms and
venograms) prior to vascular access constructions Adjuvant therapies to improve vascular access patency
(anticoagulants, antiplatelet therapies, etc.) Unknown value of new nonautogenous AVF conduits
(modified Dacron, homografts, etc.)
Role of duplex surveillance in identifying the failing graft
