Creating EndoAVF: New Evidence and Upcoming Trials

An exploration into the ins and outs of endovascular arteriovenous fistulae (endoAVF) creation was laid bare this morning during a session dedicated to innovative solutions for creation and maintenance of AV-access. In conversation with LINC Today, Ounali Jaffer, MD, from Barts Health NHS Trust in London, United Kingdom, walked through his experience in using endoAVF and touched on the past, present, and future insights that may potentially drive this field. 

EndoAVF is an exciting addition to vascular-access options for dialysis. How has the evidence supporting its use evolved in recent years, and what new clinical data are most compelling? 

As is the case with any new technology, the evidence base grows as the number of centers which take up the technology increases. Initial Industry sponsored trials showed very favorable outcomes in terms of technical success, maturation and cumulative patency rates for both devices.^1,2 We are now seeing similar outcomes in a host of independent studies from various centers worldwide. 

Within the last year, perhaps the most compelling publication was by Klein et al. in CVIR. The study was a retrospective analysis of 112 patients who had fistulae created using the 4F WavelinQ device (BD) across 3 centers in Canada, USA, and the UK. Technical success was 97.3%, with a functional maturation rate of 87% and a cumulative 2-year patency of 91.7 %.³ This compares favorably with previous studies using the Ellipsys device, such as that by Beathard et al.,⁴ which demonstrated a 2-year cumulative patency rate of 92.7%. Taken together, the data suggests that endoAVF creation using either device is a durable technique and should merit consideration for patients with chronic renal disease. 

How does the efficacy and safety profile of endoAVF compare to that of surgical AVF, particularly in terms of maturation rates, patency, and the need for reintervention?

Surgical fistulae have been shown to require approximately 4 months to mature, with a primary failure rate of 23%, a cumulative 2-year patency of 63% and a reintervention rate of 1.9 per patient year. Therefore, on the face of it, endoAVF appears more favorable.^5-8 In reality, the argument is more nuanced, as the site of access creation can have a bearing on results, as can the experience, infrastructure and volumes of procedures performed in individual centers. Although various publications have looked into this, to date there are no randomized controlled trials (RCTs) to facilitate true comparison.^9-11 

Having said this, in my view, endoAVF should be considered as an ancillary option for fistula creation as opposed to a direct competitor to surgical AVF. 

What are the critical considerations for determining patient eligibility for endoAVF creation, and how do you navigate cases with borderline anatomy?

Before considering endoAVF, you need to ensure that patients are readily available to attend all follow-up appointments because, if reintervention is required, it must be done in a timely manner. Patient eligibility then largely depends on the patient’s anatomy. In my experience, anatomical feasibility for endoAVF is approximately 50–60% of all patients screened. Personally, I feel if you start venturing off instructions for use the chances of success start to diminish. If anatomy appears borderline, then perhaps consider other options first. 

As you mentioned, the WavelinQ and Ellipsys systems are the two primary devices for creating endoAVF. What are the key differences between them, and how do you decide which device to use? 

Whilst both devices create an endovascular fistula within the upper forearm, they are in reality completely different devices. WavelinQ creates the fistula within the deep venous system, close to the perforator, between either the ulnar or radial artery and one of its paired veins. The Ellipsys fistula is between the proximal radial artery and the perforator/deep communicating vein. Aside from this, the sites for access, the AVF creation method, and the image guidance requirements are also completely different. 

When it comes to device selection, a lot resides on the anatomical screening but if both options are available, I would usually start with WavelinQ device, as an Ellipsys fistula can still be subsequently formed if required, but the same is not true vice-versa. Procedure times are generally quicker for the Ellipsys compared to WavelinQ. 

Current studies highlight promising short- and mid-term outcomes for endoAVF. What do we know about long-term durability, and what research gaps still need addressing?

In reality, we know very little beyond the 2-year mark. The Ellipsys pivotal study reported a very respectable 5-year cumulative patency rate of 82%.¹ WavelinQ data I am sure will follow, but we have to be mindful that the system was changed from a 6 F to a 4 F device, so there is an understandable lag compared to the Ellipsys system. Perhaps the one thing that is really missing is an RCT comparing the 2 devices. 

How do you see endoAVF reshaping the landscape of vascular access for dialysis patients?  Do you anticipate it becoming a first-line option in specific patient populations?

AVFs are lifelines for patients with end-stage kidney disease, and we know there are limited anatomical options for surgical AVF creation, so anything that allows us to expand the potential real estate available to create these lifelines seems like a good idea. The algorithm proposed by Robert Shahverdyan advocating a distal to proximal approach is what we follow in our institution.¹² 

We begin by assessing suitability for a radiocephalic fistula (RCF) and then move into the proximal forearm if this is not an option. I suspect that as techniques and devices evolve—and if the body of evidence expands and remains favorable—the scope and utilization of endoAVF will widen. Currently, for us, the device can be potentially used first line if an RCF is not an option and/or it is the express wish of the patient. In these groups, cosmesis is often cited as a reason for choosing endoAVF over surgical AVF. 

Can you share details about any upcoming trials focused on endoAVF, and what key questions these trials are aiming to answer?

Both WavelinQ and Ellipsys are currently recruiting for their post-market surveillance studies in the USA. These datasets should be useful additions to the growing evidence base. Perhaps the most interesting will be two RCTs comparing endoAVF with surgical AVF. One of these is being conducted by the group from University of California, Los Angeles in the USA. Patient recruitment has already begun, with a target recruitment of 90 patients and a projected study completion date of mid-to-late 2027. The primary outcome measure for this trial is physiological maturation rates at 6 months. 

The endoAVF study due to commence in France has also generated a lot of interest. The primary outcome measure for this study is the cost–utility differential with endoAVF vs open surgery, which I think is a very pertinent issue that needs addressing within a RCT setting. The study also has a wide ranging and fairly comprehensive set of secondary outcome measures. 

The successful delivery of endoAVF requires collaboration across specialties. What lessons can you share about building a multidisciplinary framework to support this service? 

From my own personal point of view, an effective multidisciplinary team (MDT) approach is crucial to the success of an endoAVF program. To achieve this, you first need to ensure all key stakeholders are engaged and consulted early on in the process to ensure all relevant views are heard and considered. Each stakeholder then must be made aware of their defined roles in delivering the service. This then allows for a robust framework to build an MDT for shared decision-making.

The MDT within our institution includes nephrologists, vascular-access surgeons, interventional radiologists, interventional radiology clinical nurse specialists (IR CNS), vascular scientists, anesthetists, and dialysis nurses. Together we are able to formulate the patient’s dialysis life plan, safeguard appropriate patient selection, list (and timely so) the patient for their procedures, re-discuss cases of non-maturation and dysfunction, and plan the next steps in the patient’s dialysis life plan in cases of endoAVF failure. In the event of patient complications or suspected fistula malfunction, early rediscussion in the MDT through efficient feedback loops is essential. This should ideally be achieved through a central coordinator, who at our institution are the IR CNS.¹³ 

Looking ahead, how do you see the field of endoAVF evolving? Are there any new technologies, techniques, or procedural advancements on the horizon that excite you? 

As has been the case with previous technologies, there will always be continuing refinements of techniques and evolution of endoAVF devices as time progresses. We have seen this already with the addition of adjunctive procedures at the index procedure to improve outcomes and evolution of the WavelinQ device. Talk now revolves around issues such as whether new sites for fistula formation using these devices within the arm should be considered, and whether drug-coated balloons have a role. Perhaps most interesting is the so called endoAVF 2.0 devices such as Velocity. Early results from the Venos-1 feasibility study appear to be quite promising, with the US-based Venos-2 study due to start recruiting this month. 

References

1. Hull JE, Jennings WC, Cooper RI, et al. Longterm results from the pivotal multicenter trial of ultrasound-guided percutaneous arteriovenous fistula creation for hemodialysis access. J Vasc Interv Radiol. 2022;33(10):1143-1150. 

2. Lok CE, Rajan DK, Clement J, et al. Endovascular proximal forearm arteriovenous fistula for hemodialysis access: results of the prospective, multicenter novel endovascular access trial (NEAT). Am J Kidney Dis. 2017;70(4):486-497.

3. Klein E, Repko B, Alvarez A, et al. Multi-center two-year patency outcomes of endovascular arteriovenous fistulas (endoAVF) created with a 4 French system. Cardiovasc Intervent Radiol. 2024;47(8):1045–1054. 

4. Beathard GA, Litchfield T, Jennings WC. Two-year cumulative patency of endovascular arteriovenous fistula. J Vasc Access. 2020;21(3):350–356. 

5. Lok CE, Huber TS, Orchanian-Cheff A, et al. Arteriovenous access for hemodialysis: a review. JAMA. 2024;331(15):1307–1317. 

6. Al-Jaishi AA, Oliver MJ, Thomas SM, et al. Patency rates of the arteriovenous fistula for hemodialysis: a systematic review and meta-analysis. Am J Kidney Dis. 2014;63(3):464–478. 

7. Bylsma LC, Gage SM, Reichert H, et al. Arteriovenous fistulae for haemodialysis: a systematic review and meta-analysis of efficacy and safety outcomes. Eur J Vasc Endovasc Surg. 2017;54(4):513–522. 

8. Rajan DK, Ebner A, Desai SB, et al. Percutaneous creation of an arteriovenous fistula for hemodialysis access. J Vasc Interv Radiol. 2015;26(4):484–490. 

9. Inston N, Khawaja A, Tullett K, et al. WavelinQ created arteriovenous fistulas versus surgical radiocephalic arteriovenous fistulas: a single-centre observational study. J Vasc Access. 2020;21(5):646–651. 

10. Harika G, Mallios A, Allouache M, et al. Comparison of surgical versus percutaneously created arteriovenous hemodialysis fistulas. J Vasc Surg. 2021;74(1):209–216. 

11. Shahverdyan R, Beathard G, Mushtaq N, et al. Comparison of Ellipsys percutaneous and proximal forearm Gracz-type surgical arteriovenous fistulas. Am J Kidney Dis. 2021;78(4):520–529.e1. 

12. Shahverdyan R, Beathard G, Mushtaq N, et al. Comparison of outcomes of percutaneous arteriovenous fistulae creation by Ellipsys and WavelinQ devices. J Vasc Interv Radiol. 2020;31(9):1365–1372. 

13. Lie G, Ahmed N, Shah N, et al. Adapting a dialysis service for delivery of percutaneous arteriovenous fistulas. Radiographics. 2022;42(6):1795–1811.