Response to Letter by Bonassi on Article, “Risk of Death Following Application of Paclitaxel‐Coated Balloons and Stents in the Femoropopliteal Artery of the Leg: A Systematic Review and Meta‐Analysis of Randomized Controlled Trials”

To the Editor:

We thank Dr Bonassi for his letter on our recently published article in the Journal of the American Heart Association (JAHA) demonstrating a significantly increased risk of death with the use of paclitaxel‐coated balloons and stents in the femoropopliteal artery of the lower limbs.1 Dr Bonassi notes that selective dropout of studies beyond the first year of follow‐up may explain the significantly increased long‐term risk of death, and he argues that this fact may have introduced a systematic error and drive nonrandom publication bias.

The authors herein present more information about the length of follow‐up period available in relationship to the date of the original registration of individual randomized studies, and they also take the opportunity to update the meta‐analysis with more evidence that has emerged since our original publication for the sake of completeness.2 At 1 year, we analyzed 28 studies with registration years ranging from 2005 to 2017. At 2 years, we analyzed 16 studies with registration years ranging from 2005 to 2013. Only 4 studies reported 5‐year follow‐up and were registered in years 2005 to 2011. There was a significant linear correlation between the registration year and the length of the reported follow‐up period (R ²=0.34, P=0.0011).

In all, 16 studies with 2723 patients reported the incidence of all‐cause patient death at 2 years. Calculated risk ratio was 1.48 (95% CI, 1.06–2.08), absolute risk difference was 3.0% (95% CI, 1.2%–4.7%), and there was no evidence of systematic publication bias (updated forest and funnel plots available in Figure S1). At 5 years, 4 studies reported all‐cause mortality for a total of 1340 cases. Calculated risk ratio was 1.62 (95% CI, 1.20–2.17), absolute risk difference was 6.4% (95% CI, 2.7%–10.1%), and there was no evidence of systematic publication bias (updated forest and funnel plots available in Figure S2).

Clearly, the length of the follow‐up period available for analysis depended on the chronological order of the execution of the studies, as would have been anticipated by the prospective nature of randomized controlled trials. In addition, we have presented an updated meta‐analysis at 2 and 5 years with an increased sample of a predominantly claudicant population, which again confirms a significantly increased risk of death with the use of paclitaxel‐coated devices in the femoropopliteal artery. Of special note, at 2 years, 13 of the 16 studies reported a risk ratio >1. Moreover, at 5 years, all 4 studies reported a risk ratio >1. To date, however, a specific causal mechanism of the increased mortality remains unknown and few trials have collected long‐term follow‐up data.

Disclosures

None.

Supporting information

Figure S1. Random effects forest plot of all‐cause death at 2 years. Pooled point estimate was expressed as risk ratio (RR; top) and as risk difference (RD; bottom).

Figure S2. Random effects forest plot of all‐cause death at 5 years. Pooled point estimate was expressed as risk ratio (RR; top) and as risk difference (RD; bottom).

J Am Heart Assoc. 2019;8:e012172 DOI: 10.1161/JAHA.119.012172.