Skip to main content

Paclitaxel Comes Up Clean in Multiple Real-World Datasets

Paclitaxel Comes Up Clean in Multiple Real-World Datasets

Author Information:

Debra L. Beck


Hollywood, FL (January 23, 2020)—Findings from 7 different real-world data sources show no difference in survival following treatment with drug-coated and non-drug-coated devices, according to a presentation given today at the International Symposium on Endovascular Therapy (ISET).

Eric Secemsky, MD, Director of Vascular Intervention at Beth Israel Deaconess Medical Center in Boston, Massachusetts, presented to ISET attendees a wealth of both published and unpublished data on drug-coated devices in the treatment of femoropopliteal disease.

“We really need to supplement the trial data with our data,” said Secemsky. Although he acknowledged that randomized trials remain the gold standard for determining safety and efficacy, “the strength of these analyses is not only in the fact that it's a generalizable population that we lose without clinical trials, but we also have the power of numbers and we have longer term follow-up.”

Illustrating this poignantly, Secemsky showed a pie chart of the available data on DCB patients, only a small sliver of which (<2,200 subjects) were considered in the pooled meta-analysis by Katsanos et al, compared to the approximately 180,000 subjects included in Secemsky’s analyses. 

Secemsky’s real-world, patient-level data offers a counterpoint to a 2018 meta-analysis that showed an increase in all-cause mortality with paclitaxel stents and balloons when used to treat femoropopliteal disease.1

His analyses included 3 separate deep dives into Centers for Medicare and Medicaid (CMS) data, the last of which was done at the behest of the United States Food and Drug Administration (FDA) panel assessing the safety of paclitaxel-eluting devices in endovascular intervention. That analysis included 152,473 Medicare beneficiaries who underwent femoropopliteal revascularization between January 2015 and December 2017 at 3042 U.S. institutions.

The cohort was a relatively healthy population with a substantial number of claudicants and compared treatment with drug-coated devices (stents or balloons) to non-drug-coated devices (bare metal stents or percutaneous transluminal angioplasty [PTA]).

Median follow-up was 799 days and showed an advantage for drug-coated devices for the mortality endpoint in an unweighted analysis (43.1% for drug-coated vs 47.6% for non-drug-coated; hazard ratio [HR], 0.84; P<.001). This difference remained, but attenuated after adjustment for patient and hospital factors (44.7% vs. 46.1%; HR, 0.94; P<.001).

“We see [in both analyses] really no evidence of harm with drug-coated devices compared to non-drug-coated devices through more than four years of follow-up for some patients, so we really have not been able to replicate the signal of harm seen in the Katsanos meta-analysis in a large population of real-world patients treated with these devices,” said Dr. Secemsky.

Even when the researchers created an RCT-like population of more than 15,000 real-world patients, followed for a median of over 2.5 years, no different was seen in mortality (HR, 0.87; P=.004).

“We were asked to comment on the fact that maybe patients were dying more frequently in the Medicare database because they were older and less healthy, and so they died before we saw the harm of paclitaxel…but we saw no evidence of harm even within this trial-like population,” said Secemsky.

Secemsky’s presentation included two previously unpublished datasets, both of which will be published shortly. In the first, the OPTUM cohort was presented to the FDA in June 2019 and showed, once again, no difference in the cumulative incidence of all-cause mortality between drug- and non-drug-eluting devices (HR, 1.09; P=0.11) with follow-up out to a median of 763 days.

OPTUM is a mixture of private insurer data and Medicare Advantage. Medicare Advantage tends to represent typically a healthier Medicare population, said Secemsky.

No difference in survival was seen in both unadjusted and adjusted analysis, in subgroups with and without critical limb ischemia (CLI), in inpatients and outpatients, and in those who received drug-eluting stents (vs bare-metal stents), and those treated with drug-eluting balloons (vs PTA).

Asked how he manages the FDA advice to consider reserving drug-coated devices for sicker patients, Dr. Secemsky said this:

“We treated everyone with drug-coated devices before the meta-analysis was published. I treat a very sick population, primarily CLI, but also sick claudicants, and it's hard to separate who's the ‘healthy of the sick’, so I try not to do that…these are sick patients and the few times that I've put in a non-drug-coated device in the past year, I've been burnt pretty quickly. It reminds me of why, when we were looking at these data that are remarkable that we have 70 or even 80% at two and four years of patency, [and with low] clinically-driven target lesion revascularization (TLR), it shows we've done something dramatically different for this disease process, and it’s been taken away this past year on data that we don't really understand.”


  1. Katsanos K, Spiliopoulos S, Kitrou P, Krokidis M, Karnabatidis D. 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. J Am Heart Assoc. 2018 Dec 18; 7(24): e011245. doi: 10.1161/JAHA.118.011245.
Back to Top