A functioning dialysis vascular access is critical to the delivery of lifesaving hemodialysis. Arteriovenous graft (AVG) is a surgically created vascular access used for hemodialysis in patients with end-stage renal disease. AVG thrombosis due to underlying flow-limiting stenosis of the graft vein junction and outflow veins is a common complication. Thrombosed AVG can be treated with thrombolysis combined with percutaneous transluminal angioplasty with good immediate success rates. However, the mid-to-long term patency rates following angioplasty have been suboptimal. Sirolimus drug-coated balloon has been shown to be safe and effective in the salvage of thrombosed arteriovenous graft. The investigators hypothesize that sirolimus drug-coated balloon is non-inferior to stent graft in maintaining the patency of thrombosed AVG that is successfully salvaged. This study is conducted to compare the 6-month access circuit primary patency of thrombosed AVG treated with sirolimus drug-coated balloon versus stent graft.

Arteriovenous graft (AVG) is a surgically created vascular access used for hemodialysis (HD). Neointimal hyperplasia frequently occurs within AVG, resulting in flow-limiting stenosis and thrombosis. The primary patency rates of AVGs after creation were 58% at 6 months and 33% at 18 months. Hence, vascular access failure has been a common reason for hospitalization among HD patients. The culprit lesion in AVG thrombosis is usually the graft-vein (GV) junction stenosis. Histopathology examination of thrombosed AVG demonstrated maximal neointimal hyperplasia at the GV junction. Thrombolysis combined with percutaneous transluminal angioplasty (PTA) of underlying stenosis has been the standard of care to restore the function of thrombosed AVGs. Despite its widespread availability and minimally invasive nature, the mid- and long-term patency of PTA for thrombosed AVG is far from satisfactory. The reported access circuit primary patency (ACPP) rate of thrombosed AVG following thrombolysis combined with PTA alone was only 23.5% at 6-month. This is because PTA is purely a mechanical intervention that does not address the underlying biological process of neointimal hyperplasia.

Stent graft is a type of vascular stent with expanded polytetrafluoroethylene (ePTFE) coating. It provides mechanical scaffolding that keeps vascular lumen open while the ePTFE coating limits neointimal hyperplasia. The therapeutic efficacy of stent graft in the treatment of graft vein junction stenosis has been convincingly demonstrated in several randomized controlled trials (RCTs). For thrombosed AVG, the reported access circuit primary patency rates at 6-month following stenting were 34% compared to 21.8% with PTA alone. The National Kidney Foundation Kidney Disease Outcomes Quality Initiative (KDOQI) Clinical Practice Guideline for Vascular Access: 2019 Update therefore suggests the use of stent-grafts in preference to PTA alone to treat clinically significant GV junction stenosis in AVG for a better 6-month postintervention outcome. Despite its superiority compared to PTA in maintaining the short- and long-term patency of thrombosed AVG, stent graft is expensive and may not be suitable for all patients as it could impede future surgical revision or creation of a secondary vascular access in the ipsilateral arm. Furthermore, the deployment of stent graft is technically challenging and requires refined interventional skills in experience hands to ensure accurate landing and prevent migration.

Drug-coated balloon (DCB) is a balloon catheter that is coated with an anti-proliferative drug. The drug is delivered directly to the vessel wall upon balloon inflation to prevent neointimal hyperplasia and restenosis. The efficacy of DCB have been shown in several randomized controlled trials. Liao et al demonstrated that angioplasty with DCB resulted in a modest improvement in primary patency of AVG with GV junction stenosis. We conducted a prior pilot study to examine the feasibility of Sirolimus DCB angioplasty at the GV junction following successful PTA for thrombosed AVGs. Our study demonstrated an ACPP of 55% at 6 months. This finding could imply that Sirolimus DCB can be an alternative option to stent graft deployment which is the recommended treatment. DCBs are cheaper and will not impede future secondary vascular access creation. Furthermore, DCB angioplasty is similar to PTA with plain balloon procedurally, thus it is not as technically challenging as stent graft deployment and does not require highly skilled operators.

The investigators aim to perform a single-centre 2-arm parallel randomized non-inferiority trial to compare the 6-month post-intervention access circuit primary patency (ACPP) of sirolimus DCB versus stent graft in the management of thrombosed AVG.

RECRUITMENT Patients who are admitted to the hospital for thrombosed AVG will be screened and offered enrollment if they meet the inclusion criteria. Patients will be consented if they fulfill the preliminary eligibility criteria and agree to participate in the study. Patients who have successful thrombectomy of the AVG is entered to the trial and randomized to receive either sirolimus DCB or stent graft. Patients who are unsuitable for the trial will be treated in the according to hospital standard protocol.

STUDY PROCEDURE Thrombolysis of the AVG will be performed as per standard hospital protocol. Briefly, an antegrade and retrograde vascular sheath will be placed in the thrombosed AVG under ultrasound guidance. Thrombolytic agents will be delivered into the thrombosed AVG via the vascular sheaths or catheters. The culprit venous stenosis will be treated using plain angioplasty balloon and thrombus within the AVG macerated using the same angioplasty balloon. If needed, the arterial plug will be dislodged by a 5.5Fr Fogarty balloon. Following successful thrombolysis of the AVG, the number and location of lesions within the dialysis circuit will be documented. After successful thrombolysis and adequate angioplasty of the GV junction (defined as <30% residual stenosis after treatment at rated burst pressure of the angioplasty balloon), patient will be randomized to either receive sirolimus DCB or stent graft deployment at the GV junction. To ensure adequate contact, the DCB or stent graft should be of the same or oversized by 1mm the diameter of the AVG. The DCB will be inflated up to the manufacturer's stated burst pressure. An inflation device with a pressure gauge will be used to inflate up to manufacturers' stated burst pressure for a duration of 2 minutes. For patients who are randomized to stent graft, a same diameter plain angioplasty balloon will be used to fully expand the stent graft after deployment. Post treatment with DCB or stent graft, a completion graftogram will be done to document the results. The size of the balloon/stent used, inflation pressure, number of inflations, procedure complications (if any), and residual stenosis will be recorded.

POST-PROCEDURE REVIEW AND FOLLOW-UP

Immediate Post-intervention Follow-up

All participants will be assessed by physical examination for the presence of a palpable, continuous thrill as a measure of technical success. Clinical success is defined as the patient's ability to undergo at least one hemodialysis treatment at the prescribed blood flow rate for the prescribed treatment duration.

Post-intervention 1-month Follow-up

All patients will be contacted via phone at 1-month ± 2 weeks to assess for any treatment site-related adverse event within 30 days.

Post-intervention 6-month Follow-up

The participant returns for a 6-month ± 4 weeks follow-up in the clinic to be assessed by the study team for the patency rate of AVG: a palpable, continuous thrill. Dialysis charts will be reviewed for clinical indicators of significant stenosis.

Post-intervention 12-month Follow-up

Each patient will be followed-up in clinic at 12-month ± 4 weeks follow-up in the clinic to be assessed by the study team for the patency rate of AVG: a palpable, continuous thrill. Dialysis charts will be reviewed for clinical indicators of significant stenosis.

Reintervention Participants who required repeat thrombolysis of the AVG are considered to have reached primary endpoint and deemed to have completed the study and will not be replaced. The study team will continue tracking their progress until the end of intended 12-month study participation. The tracking includes review of their medical records, and if necessary, telephone call to the patient, to complete the data collection.

Expected duration of study The total duration of study for each patient upon randomization will be 12-months. The study ends after the 12-month clinic follow-up.

*Patency definitions are defined based on SIR reporting standards (Gray et al., 2003):

Post-intervention target lesion patency:

Interval after intervention until the next re-intervention at graft vein junction or until the access is abandoned. Percutaneous or surgical treatments of a new arterial or venous outflow stenosis/occlusion (including access thrombosis) that do not involve or exclude the graft vein junction from the access circuit are compatible with lesion patency. The creation of new access that incorporates the graft vein junction into the new access circuit is also compatible with target lesion patency.

Post-intervention access circuit primary patency:

Interval following intervention until the next access thrombosis or repeated intervention. It ends with the treatment of a lesion anywhere within the access circuit, from the arterial inflow to the superior vena cava-right atrial junction.

Post-intervention access circuit assisted primary patency:

Interval after intervention until access thrombosis or surgical intervention that excludes the treated lesion from the access circuit. Percutaneous treatments of either restenosis/occlusion of the previously treated lesion or a new arterial or venous outflow stenosis/occlusion (excluding access thrombosis) are compatible with assisted primary patency.

Post-intervention access circuit secondary patency:

Interval after intervention until the access is surgically declotted, revised or abandoned. Thrombolysis and percutaneous thrombectomy are compatible with secondary patency.

^Complications will be categorized according to SIR definitions of minor or major complications (Aruny et al., 2003):

A major complication is defined as one that:

1. require therapy, minor hospitalization (< 48 hours),
2. require major therapy, unplanned increase in the level of care, prolonged hospitalization (>48 hours),
3. leads to permanent adverse sequelae, or
4. death

A minor complication is one that:

1. requires no therapy with no consequence,
2. requires nominal therapy with no consequence; includes overnight admission for observation only.

Inclusion Criteria:

• Thrombosed AVG in the upper arm
• Successful thrombolysis of the thrombosed AVG, defined as the re-established of flow on Digital Subtraction Angiography (DSA) and restoration of thrill in the AVG on clinical examination (to be determined during procedure)

Exclusion Criteria:

• Patient unable to provide informed consent
• Previous bare metal stent or stent-graft placement within the dialysis access
• Previous treatment with DCB within 3 months
• Presence of central vein stenosis which cannot be adequately treated (residual recoil of more than 30%)
• Failure to re-establish blood flow
• Failure to adequately treat the GV junction (defined as residual stenosis of more than 30%)
• Sepsis or active infection
• Recent intracranial bleed or gastrointestinal bleed within the past 12 months
• Allergy to iodinated contrast media, anti-platelet drugs, heparin or sirolimus
• Pregnancy
• Life expectancy < 12 months based on physician's estimate (eg. active malignancy)