A randomized controlled trial of drug-coated balloon angioplasty in venous anastomotic stenosis of dialysis arteriovenous grafts.

Objective: Paclitaxel-coated balloons are used to reduce neointimal hyperplasia in native arteriovenous (AV) fistulas. However, no study specifically evaluated their effect on venous anastomotic stenosis of dialysis grafts. We aimed to compare the efficacy of angioplasty with drug-coated balloons (DCBs) and angioplasty with conventional balloons (CBs) for venous anastomotic stenosis in dysfunctional AV grafts.

Methods: In this investigator-initiated, single-center, single-blinded, prospective randomized controlled trial, we randomly assigned 44 patients who had venous anastomotic stenosis to undergo angioplasty with DCBs (n = 22) or CBs (n = 22) from July 2015 to August 2018. Access function was observed per the hemodialysis center's protocols; ancillary angiographic follow-up was performed every 2 months for 1 year after the interventions. The primary end point was target lesion primary patency at 6 months. Secondary outcomes included anatomic and clinical success after angioplasty, circuit primary patency at 6 months and 1 year, and target lesion primary patency at 1 year.

Results: At 6 months, target lesion primary patency in the DCB group was significantly greater than that in the CB group (41% vs 9%; hazard ratio [HR], 0.393; 95% confidence interval [CI], 0.194-0.795; P = .006), as was the primary patency of the entire access circuit (36% vs 9%; HR, 0.436; 95% CI, 0.218-0.870; P = .013). At 1 year, the target lesion primary patency in the DCB group remained greater than that in the CB group (23% vs 9%; HR, 0.477; 95% CI, 0.243-0.933; P = .019) but not the primary patency of the access circuit (14% vs 9%; HR, 0.552; 95% CI, 0.288-1.059; P = .056). No difference in anatomic or clinical success was observed; no major complications were noted.

Conclusions: Angioplasty with DCBs showed a modest improvement in primary patency of venous anastomotic stenosis and all dialysis AV grafts at 6 months. The short-term benefit was not durable to 1 year, and reinterventions were eventually needed.