The controversy regarding the safety of treating peripheral artery disease (PAD) with paclitaxel-coated devices has only deepened in the new year, with two recent studies suggesting opposite safety findings.
The debate began with a 2018 meta-analysis showing a late mortality signal associated with paclitaxel drug-coated balloons (DCBs) that sent reverberations through the interventional cardiology community (J Am Heart Assoc. 2018 Dec 18;7[24]:e011245).
Now, in a new meta-analysis involving eight randomized controlled trials (RCTs) and more than 1,400 patients with critical limb ischemia (CLI), the same researchers found significantly more early amputations and deaths in those treated with DCB below the knee, compared with conventional balloon angioplasty.
“The findings of our latest report add to previous evidence underpinning major safety concerns around use of paclitaxel in lower limb angioplasties – increased long-term patient mortality in cases of intermittent claudication,” lead author Konstantinos Katsanos MD, MSc, PhD, Patras University Hospital, Greece, said in an interview.
By contrast, a retrospective study of insurance claims in Germany showed no heightened mortality with paclitaxel-coated balloons and stents, compared with uncoated devices, in close to 38,000 patients with PAD.
On the contrary, use of paclitaxel-coated devices was associated with higher long-term survival, better amputation-free survival (AFS), and lower rates of major cardiovascular events in the treatment of chronic limb-threatening ischemia (CLTI).
These findings “emphasize the difference between population-based evidence and randomized trials,” lead author Christian-Alexander Behrendt, MD, University Medical Center Hamburg-Eppendorf, Germany, said in an interview.
Downstream “showers”
In the new meta-analysis led by Dr. Katsanos, published online Jan. 15, the 1,420 patients were treated with five different DCBs and 97% had CLI (J Vasc Intervent Radiol 2020 Feb;31[2]:202-12).
In up to 1-year follow-up, the paclitaxel DCB group had fewer target lesion revascularizations (TLR) than those of the uncoated device group (11.8% vs. 25.6%; risk ratio, 0.53; 95% confidence interval, 0.35-0.81) but worse AFS (13.7% vs. 9.4%; hazard ratio [HR], 1.52; 95% CI, 1.12-2.07).
The latter finding was driven by nonsignificant increased risks for all-cause death (odds ratio [OR], 1.39; 95% CI, 0.94-2.07) and major amputations (OR, 1.63; 95% CI, 0.92-2.90).
In dose-subgroup analyses, AFS was significantly worse in cases with high-dose (3.0-3.5 mcg/mm2) devices, but not in the single trial with a low-dose DCB (2.0 mcg/mm2).
“Considering the well-described downstream ‘showers’ of paclitaxel particles with current drug-coated balloons, we hypothesize that nontarget paclitaxel embolization is a plausible mechanism for distal foot and systemic toxicity,” Dr. Katsanos said.
Short time frame
Eric Secemsky, MD, of Harvard Medical School, and director of vascular intervention at Beth Israel Deaconess Medical Center, Boston, suggested in an interview that this theorized mechanism of harm in below-the-knee procedures could potentially shed light on a similar mechanism at play in above-the-knee procedures.
“We didn’t understand why people could potentially be dying in above-the-knee [procedures], and the suggestion here is that these devices might perhaps be causing particular embolization or maybe delayed wound healing,” Dr. Secemsky speculated.
However, “I don’t know that this is true, so I am cautious to say this is true,” he emphasized.
Dr. Secemsky said a strength of the Katsanos analysis is that the RCTs included more than 1,000 patients, but noted that it is hard to vet the quality and rigor of the data, as some of the studies have not yet been published. He also noted that paclitaxel-coated devices are not approved by the Food and Drug Administration in the United States for below-the-knee procedures.
Moreover, he continued, “two studies were driving the signal of harm: the IN.PACT DEEP, which included an iteration of their DCB that is no longer being tested; and the unpublished SINGA-PACLI trial. Those studies contributed most of the adverse events seen in this meta-analysis.”
In addition, the trials had different lengths of follow-up (6-12 months), he said. “Thus, the five trials with data available to 12 months are driving the 1-year findings, whereas three RCTs, including the primary RCT showing safety [Lutonix-BTK trial], only contribute data to 6 months.”
For this reason, “we are not too excited about this meta-analysis as of now, [because] all it tells us is that we need more data to support the safety of drug-coated devices in this population,” Dr. Secemsky said.
Dr. Katsanos explained that, “to address the differences in follow-up period and number of cases lost to follow-up, the primary endpoint was calculated on the log-hazard scale and expressed as a hazard ratio, as recommended for time-to-event outcomes.”
He highlighted that a short-term time frame of 6 months to 1 year was chosen “because it is clinically relevant to limb-threatening CLI.”
Sensitivity tests also “showed consistent direction and magnitude of the summary treatment effects in case of both AFS and freedom from TLR,” Dr. Katsanos emphasized.