Survival and Complications of Arteriovenous Fistula Dialysis Access in an Elderly Population

Introduction

The creation and survival of an arteriovenous fistula (AVF) for hemodialysis is influenced by the comorbidity present at the time dialysis is commenced. In recent years, there has been a worldwide increase in the number of elderly uremic patients requiring renal replacement therapy. A high number of pathologies, such as diabetes, arteriosclerosis and heart failure, are present in this category of patients (1). By the end of the 1990s, the number of patients aged over 75 entering dialysis exceeded the number of those aged between 65 and 75. Currently, new patients over 75 make up 35% to 45% of the totals in national and European statistics, with the prevalence rate in the 30% to 35% range (2). The overall clinical condition of patients over 75 is not very different from patients between 65 and 75 whose survival rate in dialysis treatment ranges between one and three years, depending on the study, and is equivalent to the two year average in European records (3–4). Procedures and associated complications involved in creating vascular access constitute an important cause of morbidity and hospitalization. This highlights the importance of careful and thorough assessment of the patient prior to selecting the type of access and the timing of its creation (5). In addition, the native AVF, followed by prosthetic vascular access, currently represents the first choice for vascular access. Tunneled central venous catheters (TCVC) must be considered a third choice because of the greater incidence of thrombotic and infection-related complications (6).

AIM

A retrospective analysis was performed in order to assess the survival rate and incidence of complications in AVF created in elderly patients aged 75 or older at the Nephrology Unit in Verbania.

Materials and Methods

In the period between January 2000 and August 2008, 89 patients aged 75 or over, with a mean age of 82.5±7.5, commenced hemodialysis. Eleven (12%) patients began treatment with a TCVC as first choice of access. The remaining 78 (46 men and 30 women) began dialysis with an AVF. Of these, 31% were diabetic and 46% were under pharmacologic therapy for ischemic heart disease, while all had a LVEF ≥30%. For this group of patients, 90 AVF were created: 41 distal AVF (45%), 32 proximal AVF (35%), 13 midarm AVF (15%) and four AV grafts (5%) (Fig. 1). The survival of the accesses was observed until August 2010.

OPEN IN VIEWER

Fig. 1

AVF types.

The distal AVF were all created with radial-cephalic side-to-end anastomosis at the wrist; five of the midarm AVF were created with the radial artery to perforating or median vein side-to-end anastomosis below the elbow crease and eight with radial-median side-to-side anastomosis; the proximal AVF were all side-to-end anastomosis using the humeral artery, in 22 of the cases to the median basilic vein, in seven to the perforating vein and in six to the median cephalic vein; all four grafts were PTFE prostheses, three inserted as a loop in the forearm using the humeral artery to the basilic vein and only one in the upper third of the arm between the humeral artery and the humeral vein.

During the first three years of the study, pre-operative venography was performed in selected cases in addition to a routine clinical examination. From 2003 onwards, pre-operative vessel mapping was always conducted using color Doppler ultrasound studies to assess flowmetry values, in particular the unchanged morphology of the arterial wall and peak systolic velocity, the permeability of the vein and, above all, the luminal diameter. Cut-off values of 2 mm for the diameter of the artery and 2.5 mm for the vein were used, failing which a prosthetic graft was indicated. In this instance, the diameter of the inflow vein of the graft segment required a diameter of at least 4 mm which conforms to the findings of Silva et al who showed that such practices made it possible to increase the number of AVF created and to reduce the percentage of early failures (7). Antiplatelet therapy was administered to all patients unless there were specific contraindications, in which case 100 mg of acetylsalicylic acid (ASA) or 250 mg of ticlopidine as an alternative in case of gastric intolerance to ASA, were administered. The dual anti-aggregant regimen, ASA and ticolpidine, was reserved for prosthetic accesses, on the basis of a local anti-platelet protocol.

The vascular access was monitored by means of a weekly physical examination. A color Doppler ultrasound study was performed only when abnormalities were encountered in the physical assessment: a decreased or absent palpable thrill and bruit on auscultation, the presence of pulsations, or evident vascular access lesions. Significant abnormalities detected during Doppler ultrasound study, particularly for hemodynamically relevant stenosis with double the rate of peak systolic velocity at the site of the lesion compared to that of the segment above it, constituted criteria for performing an angiogram (8).

Results

The data, analyzed using the Kaplan-Meier method, show an access survival rate of 90% after six months, 76% after 12 months and 71% after 24 months (Fig. 2). Of the patients observed in the study, 47 (60%) died with a functioning AVF. Three patients were lost to follow-up. From an analysis of native accesses in relation to the site of their creation, a greater global survival rate of the midarm AVF emerged in comparison both to the distal and proximal AVF (90% at 12 months vs. 76% and 70%; 90% at 24 months vs. 67% and 56%) (Fig. 3; Fig. 4; Fig. 5).

OPEN IN VIEWER

Fig. 2

AVF total: 87. Probability months.

OPEN IN VIEWER

Fig. 3

Distal AVF: 40. Probability months.

OPEN IN VIEWER

Fig. 4

Proximal AVF: 31. Probability months.

OPEN IN VIEWER

Fig. 5

MID-ARM AVF: 13. Probability months.

There were 18 thromboses (20%): a new AVF was created in 13 cases and five patients were converted to TCVC. Four early failures (4.5%) with the access surviving less than three months were registered. A new AVF was constructed in two cases and TCVC were placed in two others. All patients had a functioning AVF with a volume flow rate ≥270 mL/min, except for those whose accesses failed early and only one other patient who never commenced dialysis.

Inadequate blood flow was the reason for the placement of a TCVC in only one case. A total of eight patients (9%) underwent conversion from an AVF to a TCVC. In three patients the access was dysfunctional, which required angioplasty in two cases and surgical revision in a third, with a mean survival rate of 24 months following the procedure.

No local complications, vascular steal, venous congestion or edema were noted. No new heart failure or worsened cardiac status was observed.

Discussion

Our data shows vascular access patency rates in elderly patients similar to those reported in younger groups (9).

Although our cohort is limited in size, a lower survival rate for proximal AVF is evident at 24 months. This fact, together with evidence that proximal AVF are at a greater risk of peripheral ischemia as well as right heart volume overload and eccentric left ventricular hypertrophy proportional to the volume load (10), would lead one to believe that this type of access may be less suitable for elderly patients. We wish to point out, however, that while the three groups of patients were similar in as far as diabetes mellitus and cardiac conditions are concerned, the group of patients for which proximal AVF were created included seven obese patients (22%) with a body mass index (BMI) >30, whereas obesity was virtually absent in the other two groups of patients. It appears important to underline this aspect in light of recent studies that have identified obesity as a condition capable of influencing AVF survival (11) because of the likelihood of soft-tissue compression of the venous outflow (12) and have observed that the superficialization of the vein is a means of salvaging the access in such circumstances (13).

The percentage of surgical or endovascular revision was limited. The number of early failures and AVF converted to TCVC fell within an acceptable range. Significant systemic and local complications were virtually absent.

In our opinion, two factors play a particularly significant role in determining the high survival rate of AVF in the elderly: the pre-operative color Doppler ultrasound study as recommended in the International Guidelines along with the timing of the AVF placement to be performed at least six months prior to the expected commencement of dialysis treatment in order to enable adequate maturation of the access (14). Routine use of pre-operative color Doppler ultrasound appears to be of considerable importance in increasing the placement of AVF in older patients – to identify the best site for constructing the vascular access, and therefore reduce the number of failures attributable to AVF that fail to mature because of inadequate luminal diameter or poor arterial inflow (15). Another aspect which merits attention is creation of the AVF as distal as possible considering the individual patient's vascular status, in order to minimize the incidence of related complications such as a high-flow access or vascular steal phenomenon with clinically significant peripheral ischemia (16). This explains why, especially over the past three years, we have tended to create an increasingly higher number of midarm AVF using the proximal portion of the radial artery when it is not possible to create a distal fistula at the wrist: this access appears to be associated with a good survival rate and a lower rate of complications (17).

Even for an elderly patient, the decision to create an AVF with native vessels or prosthesis is a choice backed by the currently accepted opinion that, despite the ever increasing number of TCVC in use, patients with this type of access are burdened with much higher mortality, morbidity and hospitalization rates than patients with AVF (18, 19). Furthermore, even without infections, the inflammatory state induced by the presence of the TCVC will have a negative effect on hemoglobin levels and nutritional status and therefore contribute to making the already fragile elderly patient even more vulnerable (20, 21).

Our results may be affected by the higher mortality of elderly patients, which allows for earlier death with a still-functioning access.

Conclusions

Construction of AV fistulae in an elderly population can achieve excellent results, compared to other groups of dialysis patients, with a preference for the creation of the AVF at the most distal site possible (22).

While fully agreeing that it is important to assess all risk factors, especially cardiovascular (23), when choosing the type of vascular access to be created in the elderly uremic patient we strongly believe that the advanced age of patients requiring hemodialysis treatment should not be prejudicial in attempting to provide them with the best type of vascular access which, to date, is still the AVF.