A Very High-flow Arteriovenous Fistula With Added Bleeding Risk: One Solution for Two Problems

VASCULAR DISEASE MANAGEMENT 2020;17(9):E179-E182

Abstract

An 85-year-old man with a brachio-cephalic arteriovenous fistula (AVF) with asymptomatic high flow and aneurysms, was hospitalized due to congestive heart failure, growth of aneurysms with exacerbation of skin lesions as well as hand and forearm pain. Blood flow (Qa) increased to 4.0 L/min. Surgery was proposed due to the presence of skin lesions with a high bleeding risk and inadequate blood flow in the AVF with steal syndrome. It was decided that a resection of the aneurysmal venous segment with the interposition of a native vein segment needed to be performed in the patient. Significant clinical improvement was noted by the patient and skin lesions were also resolved. Qa reduced to 1.0 L/min.

Introduction

Native arteriovenous fistula is considered the gold standard of vascular access for hemodialysis due to higher patency rates and fewer complications.¹ However, problems such as stenosis, thrombosis, aneurysm, infection, high-flow fistula, heart failure and access-related hand ischemia present in some cases.² Prevalence data are extremely scarce regarding aneurysmal dilatation in clinical practice. Diagnosis of this entity depends on the criteria used but can range between 6 and 60%.³ Heart failure could develop in 25 to 50% of hemodialysis patients.¹ Access-related hand ischemia is a less frequent but worrying complication and could reach approximately 10 to 25% of cases in proximal native fistulas.¹

High arteriovenous flow is one of the most common reasons for aneurysm formation.² It is defined by a blood flow (Qa) exceeding 1.5-2.0 L/min or a Qa/cardiac output ratio higher than 0.3.⁴ In parallel, a hyperfunctioning access may also cause cardiac overload, cardiopulmonary recirculation, steal syndrome, and recurrent venous stenosis.⁴ Some other risk factors for aneurysmal degeneration have been identified, such as adult polycystic disease (ADPKD), time on dialysis, high-flux membranes and upstream venous stenosis.^2,3 Nevertheless, diabetes and probably mineral bone disease play a protective role.³ Development of aneurysms reduces the area available for cannulation, can lead to inefficacy for hemodialysis and skin ulceration with high risk of infection and bleeding.² In relation to steal syndrome, the most important predictors are age, female gender, diabetes, peripheral artery disease, coronary artery disease, anastomotic configuration and vein compliance.¹ 

Definition of aneurysm degeneration varies among several authors; however, a common threshold is a dilatation of more than two times the normal adjacent vein segment. A few classification systems have been proposed for aneurysmal degenerations, and recently one of four shapes defined by physical examination and Doppler ultrasound has been put forward.⁴ A more consensual classification of four grades is used for steal syndrome. Grade 1 includes signs of paleness and cold without pain; Grade 2, pain with exercise and/or during hemodialysis; Grade 3, pain at rest and Grade 4, presence of ulcers or necrosis.¹

Management of aneurysms is determined by the presence of clinical signs and symptoms, skin condition and blood flow of arteriovenous fistula.⁴ The diameter of dilatation is not an indication for treatment.⁴ A tough challenge is the timing decision for surgical intervention. Treatment strategies for true aneurysms and high-flow fistulas include ligation of the access, aneurysm resection with re-anastomosis, excision with interposition of PTFE or native vein segment.⁵ In those with distal hypoperfusion ischemic syndrome, indications for surgery are mainly painful symptoms and necrotic ulcers.⁵ Surgical approach consists of fistula ligation, restriction of blood flow (banding, plicature) or changing of arterial inflow (RUDI or PAI).⁵

Case Description

We present a clinical case of an 85-year-old man. He had preserved cognitive functions but was dependent on help for daily life activities. He had significant comorbidities such as hypertension, dyslipidemia, Parkinson´s disease, bilateral amaurosis and heart failure class II of NYHA, controled by hemodialysis and anti-hypertensive drugs. He also had an end-stage renal disease due to IgA nephropathy in high-flux hemodialysis since 2008, with a brachio-cephalic arteriovenous fistula in the left arm created in the same year, and posterior development of high flow and aneurysmal formation. From 2012, Doppler ultrasound of vascular access confirmed a stable Qa measured in the brachial artery of 3.2 L/min, without symptoms. In October 2015, he was hospitalized with dyspnea, hypertension, and hypervolemia. On physical examination he was confused and lethargic. Blood pressure was 220/90 mmHg with a heart rate of 90 bpm. SO2 was of 87% (FiO2 at 21%). He had bilateral pulmonary crackles without peripheral edema. Laboratory findings showed hemoglobin of 12.5 g/dL and troponin T of 80 ng/L. Chest radiography showed bilateral congestion. Cerebral TC was normal and electrocardiogram showed no signs of myocardial ischemia. Treatment with oxygen, ultrafiltration and nitroglycerin was initiated with clinical improvement. Echocardiogram demonstrated moderate dilatation of left ventricle with concentric hypertrophy (ejection fraction of 52%), pulmonary artery pressure of 55 mmHg (sugestive of pulmonary hypertension) and cardiac output of 6.67 L/min. Vascular access evaluation noted a growth of aneurysm size in the last few weeks and worsening of small previous skin lesions with the risk of rupture and bleeding (Figure 1). The patient also referred to hand and forearm pain. Qa in the brachial artery increased to 4.0 L/min. Peak systolic velocity (PSV) in radial and ulnar arteries were 21 and 13 cm/s, respectively. A fistulogram was performed and excluded central vein stenosis. Surgical intervention was proposed due to the presence of aneurysm-associated skin lesions with high bleeding risk and inadequate blood flow of the arteriovenous fistula with mild steal syndrome (Grade 2). After a multidisciplinary discussion involving nephrology, vascular surgery, interventional radiology and nursing, it was decided to perform the resection of the aneurysmal venous segment with the interposition of basilic vein segment from the patient himself. The length between the two tips was large and about 20 cm of the native vein was used to perform a by-pass with an end-to-end anastomosis (Figure 2, Figure 3). The surgical procedure was complex but without any incident. Skin lesions were resolved by resection in the same procedure. Hand and forearm ischemic symptoms improved after intervention, with a significant reduction of Qa to 1.0 L/min, measured by Doppler ultrasound in the first week. Radial and ulnar PSV also improved to 53 and 51 cm/s, respectively. The same arteriovenous fistula was still used for hemodialysis, with a more proximal ultrasound-guided cannulation from the first day after surgery. It was possible to repeat the echocardiogram only a few weeks later, and showed a stable cardiac output. Qa of fistula was near 3.0 L/min at the same time.

Discussion

In this case, various problems of the vascular access were identified simultaneously. High flow causing aneurysmal degeneration with skin lesions, steal syndrome and probably with some impact on a heart failure event, represented a difficult task to manage. This patient presented some risk factors for aneurysm formation such as advanced age, absence of diabetes, dialysis vintage, high-flux membrane, high flow of a proximal native arteriovenous fistula.³ High flow was probably already present a few months after arteriovenous fistula creation with progressive aneurysmal formation, due to a stable Qa of 3.2 L/min 4 years after surgery. No intervention was proposed at that time because of good clinical tolerance. Mean time reported for aneurysm formation was of 4.9 years, ranging from 1 to 14 years, as shown in Sigala et al.² However, access-related hand ischemia appeared later, only in the hospitalization of October 2015, and was probably due to the further increase in blood flow until that time. This is in accordance with the Scheltinga et al review, showing a later timing onset of symptoms with native arteriovenous fistulas compared to grafts.⁶ This fact is the result of the initial maturation period and further compliance of venous capital. As well as steal syndrome, the patient also developed heart failure exacerbation. However, the patient was always asymptomatic during the course of hemodialysis and the heart failure event probably occurred due to hypertension and fluid overload, because he only improved with anti-hypertensive drugs and ultrafiltration. The development of cardiac failure seems to have an inverse relationship to the distance between heart and anastomosis.⁶ After arteriovenous fistula creation, a compensatory increase in cardiac output follows an early decrease in peripheral vascular resistance. Eccentric hypertrophy with diastolic dysfunction and pulmonary hypertension also occurs.⁵ The echocardiogram showed some of these features, including moderate dilatation of the left ventricle and pulmonary artery pressure of 55 mm Hg. Concentric hypertrophy probably arises from arterial hypertension.

The treatment strategy in this case was extremely challenging, due to the presence of many simultaneous problems: high-flow arteriovenous fistula with steal syndrome (Grade 2) and aneurysmal degenerations with associated skin lesions. Prompt intervention was needed because of the high risk of infection and bleeding from the aneurysms. Ligation of vascular access was one possible choice but had the disadvantage of requiring a tunneled catheter placement. Aneurysmorraphy was not performed due to the unsuitable quality of the vein walls. Banding procedure could also resolve steal syndrome and heart failure but would probably increase thrombosis risk. Long-segment plicature also represented a viable technique with good results in Powell et al. They showed a post-operative patency rate of 88%, but also 5.7% (2 cases) of thrombosis.⁷ A better solution for this complex case was thought to be the resection of the aneurysmal venous segment with interposition of a native vein. Some series demonstrated that more aggressive surgical revision of aneurysms appears to be an acceptable and safe management strategy to salvage dialysis access.⁸ Interposition of graft segment was rejected in this case because of higher risk of infection and to maintain an autologous reconstruction. Georgiadis et al found another significant advantage with higher patency rates after autologous reconstruction compared to graft interposition.⁹ Although the saphenous vein is usually the best choice to restrict a Qa increase, saphenous veins were absent in our patient and he showed a good cardiac condition to support a basilic vein interposition. Pasklinsky et al also reported one case of thrombosis and two anastomotic stenoses in a group of patients treated with excision and repair with saphenous vein.¹⁰ After surgical intervention, our patient improved from symptoms of hand and forearm pain. Doppler ultrasound in the first week post-surgery showed a reduction in Qa from 4.0 to 1.0 L/min and better distal perfusion, unfortunately without a parallel echocardiogram evaluation. A few weeks after surgery, Qa of fistula was near 3.0 L/min and cardiac output was similar to its pre-surgical value, without evidence of heart failure. This could be explained by the timing of echocardiogram study and probably with further dilatation of the native vein segment. In Sigala et al, a moderate dilatation of a corrected venous segment was observed, but without symptomatic recurrence.² Thus, despite indication for surgical intervention, resolution of a heart failure event in this case was more likely to be related with medical management such as ultrafiltration and anti-hypertensive therapy. 

Conclusions

It was possible to treat two main complications by salvaging a native arteriovenous fistula, maintaining its use by immediate ultrasound-guided cannulation and without any catheter placement.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. They report no conflicts of interest regarding the content herein. The authors did not receive any specific funding.

Address for Correspondence: Rui Miguel Guimarães Abreu, Avenida do Hospital Padre Américo 210, 4564-007, Guilhufe, Penafiel, Portugal. Email: ruimigabreu@gmail.com  Phone number: +35 191 266 5243

REFERENCES

1. Ibeas J, Roca-Tey R, Vallespín J, et al. Spanish Clinical Guidelines on Vascular Access for Haemodialysis. Nefrologia. 2017;37(Suppl1):1-191.

2. Sigala F, Kontis E, Sassen R, Mickley V. Autologous surgical reconstruction for true venous hemodialysis access aneurysms—techniques and results. J Vasc Access. 2014;15(5):370-375.

3. Jankovic A, Donfrid B, Adam J, et al. Arteriovenous fistula aneurysm in patients on regular hemodialysis: prevalence and risk factors. Nephron Clin Pract. 2013;124(1-2):94-98.

4. Balaz P, Björck M. True aneurysm in autologous hemodialysis fistulae: definitions, classification and indications for treatment. J Vasc Access. 2015;16(6):446-453.

5. Sequeira A, Tan T-W. Complications of a high-flow access and its management. Semin Dial. 2015;28(5):533-543.

6. Scheltinga MR, van Hoek F, Bruijninckx CMA. Time of onset in haemodialysis access-induced distal ischaemia (HAIDI) is related to the access type. Nephrol Dial Transplant. 2009; 24(10):3198-3204.

7. Powell A, Wooster M, Carrol M, et al. Long-segment plication technique for arteriovenous fistulae threatened by diffuse aneurysmal degeneration: short-term results. Ann Vasc Surg. 2015;29(6):1327-1331.

8. Furukawa H. Surgical management of vascular access related aneurysms to salvage dialysis access: case report and a systematic review of the literature. J Vasc Access. 2015;16(2):120-125.

9. Georgiadis GS, Lazarides MK, Panagoutsos SA, et al. Surgical revision of complicated false and true vascular access-related aneurysms. J Vasc Surg. 2008;47(6):1284-1291.

10. Pasklinsky G, Meisner RJ, Labropoulos N, et al. Management of true aneurysms of hemodialysis access fistulas. J Vasc Surg. 2011;53(5):1291-1297.