Flowing Fistulas: Update on Strategies to Maintain Arteriovenous Fistula Patency Authors: Maddison Taylor, Benjamin Lazarus Published: 2026 See Clinical Research on Article 106583 See Clinical Research on Article 106583 Current State of Play: Arteriovenous Fistulas Arteriovenous fistulas (AVF) are widely used for hemodialysis and have lower infection rates and greater longevity than other modalities of hemodialysis access, however, complications with maturation and maintaining patency are common.1 Approximately 40% to 50% of fistulas fail to mature without intervention, and primary patency rates range from 34% without intervention to 83% with intervention.2 Dysfunctional vascular access impacts on patients quality of life, increases procedural burden and health care costs; approximately $3 billion is spent annually for vascular access creation and maintenance.1,3 Maintaining patency of the AVF dialysis circuit is a significant challenge and often requires multiple interventions, with a mean intervention rate of 0.37 procedures/patient year.4 The Kidney Disease Outcomes Quality Initiative suggests that < 3 interventions per year to maintain patency is a reasonable target goal.5 Interventions are recommended for clinical dysfunction, such as reduced flow and elevated venous pressures associated with > 50% stenosis.4 Pre-emptive stenting of lesions in the absence of clinical indicators was not beneficial.6 Older age, poor nutritional status, high dose erythropoietin, female sex, diabetes, peripheral vascular disease, and vessel anatomy have been associated with AVF failure, stenosis and thrombosis.7 In this regard, the study by Tan et al.8 provides important insights into potential new therapies for treatment of AVF stenosis. Neointimal Hyperplasia and AVF Stenosis The mechanism of AVF stenosis formation is well described and highlights potential targets for intervention (Figure 1).6 After AVF creation the venous endothelium is exposed to arterial pressures, leading to a change in vascular smooth muscle cell phenotype. Cells adapt by reducing expression of contractile markers, extracellular matrix deposition resulting in outward venous remodelling.9 This process can become maladaptive if there is inward remodeling, known as neointimal hyperplasia (NIH) and contributes to over 60% of dysfunctional fistulas.1,9 Physical stressors such as repeated needle stick from cannulation, shear wall stress from anatomical variations, and trauma from angioplasty have been implicated in the development of NIH.4 The inflammatory milieu of kidney failure with uremia and oxidative stress leads to gene dysregulation where vascular smooth muscle cell alter into a fibrotic phenotype; with macrophage recruitment, endothelial dysfunction, and proinflammatory cytokines predisposing to a maladaptive response within the vessel wall and resultant NIH.4,6,S10) Histologically NIH is characterized by smooth muscle proliferation within the internal elastic lamina and extracellular matrix production within the intima, leading to poor compliance and reduced vessel diameter.1,3,S11Figure 1Upstream and downstream events leading to arteriovenous access stenosis. vascular smooth muscle cells (VSMCs), extracellular matrix (ECM). Adapted from DePietro and Trerotola.6 Upstream and downstream events leading to arteriovenous access stenosis. vascular smooth muscle cells (VSMCs), extracellular matrix (ECM). Adapted from DePietro and Trerotola.6 Percutaneous Intervention Percutaneous transluminal angiography (PTA) was first developed in the treatment of arteriovenous stenosis in 1982 and is still considered first line for management of dysfunctional AVFs.5,6 It involves expanding a stenotic vessel lumen through forceful dilatation to disrupt the neointimal and tunica media layer leading to outward remodelling.6 PTA is highly effective with over 90% intervention success rates at achieving the required < 30% residual stenosis for an adequate angiographic result.6 However, the durability of this intervention to maintain AVF primary circuit patency ranges from 41% to 76% at 6 months and 31% to 50% at 12 months post procedure.1,4 Venous stenoses form at sites of turbulence such as swing veins, anastomoses, cephalic arch, and frequent needling sites— they are the most common reason for fistula failure.3,6 Given the high recurrence rates at previous angioplasty sites1 further developments into high pressure balloons, cutting balloons, drug coated balloons, and medical optimization have been explored. High-Pressure and Cutting Balloons Over the last 40 years PTA technology has progressed, largely informed by advances in cardiac and peripheral vascular disease interventions. This includes the development of high and ultrahigh pressure balloons which inflate with higher pressures for resistant stenoses, as well as cutting balloons fitted with microsurgical blades for thickened, calcified intima.6 Although both high pressure balloons and cutting balloons are comparable for immediate angiographic result, cutting balloon angioplasty has had better target lesion primary patency at 6 months in 1 trial (Table 1),S12 and may be a preferred option for some specific indications, such as graft-vein anastomotic lesions.6Table 1Strategies to maintain patency of an arteriovenous fistulaStrategySummary of evidencePossible future directionsPTA interventionsHigh pressure balloonsMost stenoses can be corrected with a balloon pressure of 15–17 atm, with an increased risk of NIH and recurrence if over 20 atm is required for treatment.6 Ultrahigh pressure balloons (30–40 atm) are available but can cause barotrauma to the vessel.6 An RCT compared CBA intervention (n = 36) against HPBA (n = 35) showed both methodologies were successful at immediate stenosis correction. However, CBA outperformed HPBA in TLPP rates at 6 mos (66.4% and 39.9% respectively, P = 0.01)S12Randomized trials comparing intervention options for resistant stenoses, such as HPBA vs. CBA vs. pre-/post-conditioning with DCB therapyCutting balloonsUsed for resistant stenoses, uses micro-blades fitted to balloon. In vein-to-graft anastomosis stenoses CBA outperformed conventional PTA for patency at 1 year (63% vs. 37%, P = 0.037)S23. Multiple other RCT have not shown a persistent benefit in cutting balloons over PTA or high-pressure balloons in native AVF stenosesS24,S25Identification of specific patient parameters or lesion types which will most benefit from cutting balloonsPaclitaxel coated balloonsCommonly used to reduce NIH and restenosis given cell cytotoxicity profile. High number of cohort and RCT trials available of variable certainty evidence.Meta-analysis of 5 RCTs pooled data (n = 1107) found no statistically significant difference in TLPP at 6 mos [OR: 1.79, 95% CI: 0.66–4.90, P = 0.181]. However, there was a trend towards reduced re-intervention rates at the target lesion and when they excluded the study which largely contributed to the heterogeneity the TLPP outcome became significant in favor of PCB intervention [OR: 2.42, 95% CI: 1.29–4.55).S26In peripheral vascular disease trials, there were concerns for increased mortality rates amongst PCB treated patients.S27 Reassuringly in meta-analysis from AVF studies there was no difference in all-cause mortality at 24 mos [OR: 1.4, 95% CI: 0.83–2.45, I2: 0%].This analysis also showed that if TLPP was measured as a time to event analysis there was a significant benefit to PCB treatments with a 6 mo hazard ratio = 0.47 [95% CI: 0.33–0.69].S13Care should be taken when considering use of PCBs for immature AVFs (< 12 mos) as a cohort trial within this cohort revealed poorer outcomes compared with PTA alone.S28Randomized trials exploring outcomes based on dose, excipient, and duration of exposure. Multiple trials have subgroup analyses suggesting higher doses or longer exposure were associated with preferable outcomes presumably because of improved drug deliver to vessel wall.Sirolimus coated balloonsA meta-analysis comparing over 70,000 patients with either sirolimus eluting stents or paclitaxel eluting stents in coronary interventions showed a significantly better overall survival, MACE, and restenosis rates amongst the sirolimus group.S29 The trial by Tan et al.8 tested sirolimus coated balloons for AVF stenosis.Further studies comparing sirolimus to other interventions in the population with kidney disease.StentingStent grafts - native AVFThe RESCUE trial which compared 6 months ACPP between patients treated with stent graft (n = 132) vs. PTA alone (n = 143). Results were favorable for stent grafts with 18.6% patency vs. 4.5% patency in PTA group at 6 mos (HR: 0.62, P < 0.001) and this statistical significance held true in native fistula subgroup analysis.S30 However, this trial raised concerns given the 4.5% ACPP patency rate after PTA is far below the reported standard and may be related to operator discrepancy or how ACPP was measured leading to over-reporting of failure. At this stage PTA remains the standard of care for initial invention of venous stenosis within native fistulae because of operator experience, ease of access, and cost efficiency. There is limited RCT data to suggest that stenting is superior to PTA within the native fistula.3,S31Delineation of which stenotic lesions, fistula types and cohorts would best benefit from expanded PTFE (polytetrafluoroethylene) grafts is an important future direction of research.Ongoing trials into biologically engineered human acellular vessel implantation devices, derived from donor vSMCs (S19)Stenting of AVGs or proximal stenosisStenting has had some success in randomized trials when treating stenoses of arteriovenous graftsS32 or specifically proximal stenosisS33 within the cephalic archS34.Medications to maintain AVF patencyAntiplateletsComposite results from 2481 patients comparing the administration of anti-platelet therapy post operatively found that primary patency was higher in those receiving antiplatelets regardless of type or dose (OR: 2.28, 95% CI: 1.42–3.65).S17 Another meta-analysis of RCTs with aspirin as adjuvant therapy showed a reduced rate of graft thrombosis (RR: 0.36, P = 0.006).S35 Omega-3 therapy has also been evaluated because of its anti-inflammatory and antiproliferative properties.Prospective trials examining effect of specific antiplatelets and dosing pre and post AVF angioplasty on access outcomes, and rates of major bleeding.Omega-3 PUFAData from 833 participants found a signal for improvement in primary patency in those taking Omega-3 PUFA (RR: 0.81, CI: 0.68–0.98) but with no difference in access abandonment, dialysis usability, and all-cause mortality.S36Exploration of specific doses and formulations of omega-3 PUFA. Further investigation of other novel agents that may reduce oxidative stress and NIH.AVF, arteriovenous fistula; AVG, arteriovenous graft; CBA, cutting balloon angioplasty; HPBA, high pressure balloon angioplasty; PCB, paclitaxel coated balloons; PTA, percutaneous transluminal angioplasty; PUFA, polyunsaturated fatty acids; RCT, randomized controlled trial; TLPP, target lesion primary patency; vSCMs, vascular smooth muscle cells. Table 1Strategies to maintain patency of an arteriovenous fistulaStrategySummary of evidencePossible future directionsPTA interventionsHigh pressure balloonsMost stenoses can be corrected with a balloon pressure of 15–17 atm, with an increased risk of NIH and recurrence if over 20 atm is required for treatment.6 Ultrahigh pressure balloons (30–40 atm) are available but can cause barotrauma to the vessel.6 An RCT compared CBA intervention (n = 36) against HPBA (n = 35) showed both methodologies were successful at immediate stenosis correction. However, CBA outperformed HPBA in TLPP rates at 6 mos (66.4% and 39.9% respectively, P = 0.01)S12Randomized trials comparing intervention options for resistant stenoses, such as HPBA vs. CBA vs. pre-/post-conditioning with DCB therapyCutting balloonsUsed for resistant stenoses, uses micro-blades fitted to balloon. In vein-to-graft anastomosis stenoses CBA outperformed conventional PTA for patency at 1 year (63% vs. 37%, P = 0.037)S23. Multiple other RCT have not shown a persistent benefit in cutting balloons over PTA or high-pressure balloons in native AVF stenosesS24,S25Identification of specific patient parameters or lesion types which will most benefit from cutting balloonsPaclitaxel coated balloonsCommonly used to reduce NIH and restenosis given cell cytotoxicity profile. High number of cohort and RCT trials available of variable certainty evidence.Meta-analysis of 5 RCTs pooled data (n = 1107) found no statistically significant difference in TLPP at 6 mos [OR: 1.79, 95% CI: 0.66–4.90, P = 0.181]. However, there was a trend towards reduced re-intervention rates at the target lesion and when they excluded the study which largely contributed to the heterogeneity the TLPP outcome became significant in favor of PCB intervention [OR: 2.42, 95% CI: 1.29–4.55).S26In peripheral vascular disease trials, there were concerns for increased mortality rates amongst PCB treated patients.S27 Reassuringly in meta-analysis from AVF studies there was no difference in all-cause mortality at 24 mos [OR: 1.4, 95% CI: 0.83–2.45, I2: 0%].This analysis also showed that if TLPP was measured as a time to event analysis there was a significant benefit to PCB treatments with a 6 mo hazard ratio = 0.47 [95% CI: 0.33–0.69].S13Care should be taken when considering use of PCBs for immature AVFs (< 12 mos) as a cohort trial within this cohort revealed poorer outcomes compared with PTA alone.S28Randomized trials exploring outcomes based on dose, excipient, and duration of exposure. Multiple trials have subgroup analyses suggesting higher doses or longer exposure were associated with preferable outcomes presumably because of improved drug deliver to vessel wall.Sirolimus coated balloonsA meta-analysis comparing over 70,000 patients with either sirolimus eluting stents or paclitaxel eluting stents in coronary interventions showed a significantly better overall survival, MACE, and restenosis rates amongst the sirolimus group.S29 The trial by Tan et al.8 tested sirolimus coated balloons for AVF stenosis.Further studies comparing sirolimus to other interventions in the population with kidney disease.StentingStent grafts - native AVFThe RESCUE trial which compared 6 months ACPP between patients treated with stent graft (n = 132) vs. PTA alone (n = 143). Results were favorable for stent grafts with 18.6% patency vs. 4.5% patency in PTA group at 6 mos (HR: 0.62, P < 0.001) and this statistical significance held true in native fistula subgroup analysis.S30 However, this trial raised concerns given the 4.5% ACPP patency rate after PTA is far below the reported standard and may be related to operator discrepancy or how ACPP was measured leading to over-reporting of failure. At this stage PTA remains the standard of care for initial invention of venous stenosis within native fistulae because of operator experience, ease of access, and cost efficiency. There is limited RCT data to suggest that stenting is superior to PTA within the native fistula.3,S31Delineation of which stenotic lesions, fistula types and cohorts would best benefit from expanded PTFE (polytetrafluoroethylene) grafts is an important future direction of research.Ongoing trials into biologically engineered human acellular vessel implantation devices, derived from donor vSMCs (S19)Stenting of AVGs or proximal stenosisStenting has had some success in randomized trials when treating stenoses of arteriovenous graftsS32 or specifically proximal stenosisS33 within the cephalic archS34.Medications to maintain AVF patencyAntiplateletsComposite results from 2481 patients comparing the administration of anti-platelet therapy post operatively found that primary patency was higher in those receiving antiplatelets regardless of type or dose (OR: 2.28, 95% CI: 1.42–3.65).S17 Another meta-analysis of RCTs with aspirin as adjuvant therapy showed a reduced rate of graft thrombosis (RR: 0.36, P = 0.006).S35 Omega-3 therapy has also been evaluated because of its anti-inflammatory and antiproliferative properties.Prospective trials examining effect of specific antiplatelets and dosing pre and post AVF angioplasty on access outcomes, and rates of major bleeding.Omega-3 PUFAData from 833 participants found a signal for improvement in primary patency in those taking Omega-3 PUFA (RR: 0.81, CI: 0.68–0.98) but with no difference in access abandonment, dialysis usability, and all-cause mortality.S36Exploration of specific doses and formulations of omega-3 PUFA. Further investigation of other novel agents that may reduce oxidative stress and NIH.AVF, arteriovenous fistula; AVG, arteriovenous graft; CBA, cutting balloon angioplasty; HPBA, high pressure balloon angioplasty; PCB, paclitaxel coated balloons; PTA, percutaneous transluminal angioplasty; PUFA, polyunsaturated fatty acids; RCT, randomized controlled trial; TLPP, target lesion primary patency; vSCMs, vascular smooth muscle cells. Strategies to maintain patency of an arteriovenous fistula AVF, arteriovenous fistula; AVG, arteriovenous graft; CBA, cutting balloon angioplasty; HPBA, high pressure balloon angioplasty; PCB, paclitaxel coated balloons; PTA, percutaneous transluminal angioplasty; PUFA, polyunsaturated fatty acids; RCT, randomized controlled trial; TLPP, target lesion primary patency; vSCMs, vascular smooth muscle cells. Drug Coated Balloons Drug coated balloons (DCB) are designed to be complimentary to percutaneous transluminal angioplasty, and were first approved by the FDA for use in fistuloplasty in 2017.1 The initial PTA procedure dilates the stenoses. Subsequently the DCB is applied with varying medications, doses, lengths, durations, and excipients in order to provide a uniform local effect of drug to the vessel, without systemic side effects.S11 Paclitaxel is the most common DCB and interferes with microtubule formation, leading to cell cytotoxicity subsequently reducing rates of NIH and restenosis4; it has been studied extensively in coronary artery disease and peripheral vascular disease.S13 There are at least 7 different paclitaxel impregnated balloons available with varying lengths, excipients and doses between 2 mg/mm2 to 3.5 mg/mm2.S11 Recent meta-analyses comparing paclitaxel coated balloons (PCB) to PTA have not revealed a statistically significant difference in target lesion primary patency rates, however there are some signals for benefit to PCB in subgroup or time to event analysis (Table 1). New interventions for coronary artery disease often lead to new treatments for AVF stenosis. The majority of drug eluting stents used in coronary stents are from the -limus family and have been found to have better target lesion revascularization rates than paclitaxel stents.S14 Sirolimus is an mTOR inhibitor and functions as an antiproliferative agent, preventing smooth muscle cell proliferation after vascular injury and reduces NIH post angioplasty; this has been shown in animal models and in clinical studies.4,S15 Recently, Tan et al.8 conducted the first multi-center randomized controlled trial comparing the effect of sirolimus coated balloons (n = 83) to placebo balloons (n = 87) on access circuit primary patency, defined as freedom from clinically driven reintervention or thrombosis of the AVF during the 6 months after the index procedure. At 6 months the access circuit primary patency was higher in the sirolimus coated balloons group compared with placebo with 70.1% patency versus 56.7% patency.S16 Although patients and managing physicians were blinded to the assigned intervention in this trial, proceduralists and trial coordinators were not blinded because of visible differences in the study balloons. This has important implications for the interpretation of the results because the assessment of access circuit primary patency is partially subjective. This is also a small trial, with limited capacity to identify differences in adverse events. Drug delivery of sirolimus to the vessel wall was not demonstrated in this study, which may be an important consideration for future trials. Nonetheless, this is an important trial and does open a new avenue to advance interventions to maintain AVF patency. Future trials may evaluate the longer-term outcomes beyond 6 months, cost effectiveness of sirolimus coated balloons, and direct comparison with other angioplasty technologies. In future trials assessment of drug delivery to the vessel wall will be necessary as sirolimus is less lipophilic with reduced tissue retention compared with paclitaxel and longevity of suppression of NIH beyond 6 months has not yet been determined.S15 Stenting and Medical Therapy Stenting has had some success when treating stenoses of arteriovenous grafts and proximal stenoses within the cephalic arch, however there is limited randomized controlled trial data to suggest that stenting is superior to PTA within the native fistula.3 Retrospective studies (Table 1) have examined the role of antiplatelets and other medications that may improve fistula patency, albeit with higher risk of bias.S17 Future Directions and Other Treatment Approach Ongoing investigation into the basic science and pathophysiology of NIH and AVF stenosis will be integral to identifying new translational targets for intervention. Metabolic responses and gene regulation are both strongly implicated in the dysfunction of AVF patency and transcriptomics have revealed significant gene regulation changes to inflammatory pathways, cellular adhesion, metabolic dysregulation, and coagulation cascade activation.9, S18 On the back of advances in discovery science and technology, other novel device therapies are in development.1, S19–S21 With new technological advances, it is important to delineate the best future therapies for maintaining AVF patency through clinical trials. Comparative effectiveness trials are required to examine and identify the optimal treatment, or combination of treatments, to maintain AVF patency for individual patients. Testing different vessel preconditioning, DCB dose, contact time and excipient used will likely reveal important insights into the best use of DCB technology. Embedding such trials in national clinical registries, S22 and using more efficient trial designs, such as coordinated multicenter adaptive platform trials, would be well suited to these future studies by reducing data collection costs, simplifying recruitment, and increasing availability of long term follow up. Future trials should also include cost-effectiveness analysis into trial design and standardize outcome measures that are important to healthcare providers and patients. Disclosure All the authors declared nocompeting interests.