Data have been presented for the first time at the Paris Vascular Insights (PVI) conference (23–25 November, Paris, France) from the ABISS trial, which compared drug-coated balloon (DCB) with plain-balloon angioplasty for arteriovenous fistula (AVF) stenosis.
ABISS was a physician-designed prospective, randomised, double-blind trial that included 12 centres in France. The study enrolled 150 adult patients who had a native autologous AVF stenosis already punctured for haemodialysis. Following predilatation, patients were randomised 1:1 between DCB (Lutonix, BD) and placebo balloon application. The study was mainly funded by the French government clinical research programme (PHRC).
The primary outcome of the study was cumulated incidence of loss of primary patency of AVF at six months. The study authors, led by Raphaël Coscas (Ambroise Paré Hospital and Paris-Saclay University, Paris, France) found in an intention-to-treat analysis DCB was superior to plain-balloon angioplasty at six months but not in a statistically significant way, with a p value of 0.09. It was however statistically significant at three months (p=0.002). In a prespecified per-protocol analysis, the DCB was superior to a statistically significant degree at the intervals of three (p=0.0004), six (p=0.008) and 12 months (p=0.029) for the main outcome. There was no difference in terms of mortality or rates of adverse events between the groups. Coscas noted that the secondary outcomes, which included dialysis circuit patency, restenosis, and thrombosis, will be presented in future meetings.
The presenter drew attention to the difference between the intention-to-treat analysis results and those of the per-protocol analysis. He said that “the way data of the study were analysed will need to be discussed because, in the intention-to-treat analysis, patients lost to follow-up or who waived consent were imputated to the worst-case scenario (loss of patency if DCB arm and patency if placebo arm),” suggesting a “disadvantage” for the DCB in the intention-to-treat results. “This was less the case in the per-protocol analysis—I think we will discuss that later on”. He added that post-hoc analyses based on similar criteria of other major DCB in AVF trials are ongoing.