No Excess Mortality Risk With Paclitaxel Stents in PAD, Second Medicare Analysis Finds

A large analysis of Medicare patients shows no difference in long-term mortality between those who received a paclitaxel-eluting stent or BMS for the treatment of femoropopliteal artery disease. The study is the second real-world investigation in recent weeks to be at odds with a meta-analysis published late last year that demonstrated an excess mortality signal for paclitaxel-based stents and balloons in the periphery compared with uncoated devices.

In the latest study, the mortality rates in the Medicare group are strikingly high, owing to a large prevalence of critical limb ischemia (CLI) patients, which accounted for about 60% of the group. However, the absolute risk of death at a median of 2 years was similar whether patients did or did not have CLI and whether they did or did not receive a paclitaxel DES.

“It’s incredibly important to show what these devices are doing in each of these populations,” lead author Eric A. Secemsky, MD (Beth Israel Deaconess Medical Center, Boston, MA), told TCTMD. “At 2 years we have almost 13,000 patients in the non-CLI population and just over 13,000 in the CLI population.”

Secemsky’s group performed a prior Medicare analysis of ICD-10 codes for 16,560 patients who underwent femoropopliteal artery revascularization with paclitaxel-based balloons and stents that showed no difference in mortality through 20 months of follow-up. In order to get longer-term data, however, his group went back into the US Centers for Medicare & Medicaid Services’ database and focused on patients who received a peripheral-artery stent between late 2012 through September 2015, with cumulative incidence of death analyzed through the end of 2016. The start of the study corresponds with Food and Drug Administration approval of the first peripheral DES (Zilver PTX, Cook Medical).

Commenting on the study for TCTMD, Christopher J. White, MD (Ochsner Medical Center, New Orleans, LA), said while this second Medicare analysis “gets deeper and more appropriately into the issue of the late mortality,” it can only be considered hypothesis-generating since it is based on administrative data and ICD codes, which he noted are notoriously heterogenous and prone to errors.

In early December 2018, a meta-analysis of 28 trials of paclitaxel DES or DCBs in PAD patients led by Konstantinos Katsanos, MD, PhD (Patras University Hospital, Rion, Greece), reported a 68% relative risk increase in all-cause death in the device group compared with an uncoated device group, with a number-needed-to-harm of 29. At 5 years, the relative risk increase was 93%, with a number-needed-to-harm of 14. Since then, several studies focused on analyzing deaths in the individual trials using patient-level data, together with the two Medicare analyses have not replicated the mortality signal, making it less and less clear what, if anything, is actually going on.

Medicare, Take Two

In this second Medicare analysis, published online ahead of print today in the JACC, Secemsky and colleagues examined data on 51,456 patients with an average age of 72.8 years. Of these, 59.7% had CLI. At median follow-up of 2 years (interquartile range 1.2-3.0 years; longest 4.1 years), mortality was 51.7% for DES versus 50.1% for BMS (log-rank P = 0.16). There was also no difference in the probability of death between DES versus BMS in patients with CLI (log rank P = 0.52) or for DES versus BMS in patients without CLI (log-rank P = 0.55).

Additionally, after multivariable adjustment, no association was seen between stent type and mortality (HR 0.98 for DES vs BMS; 95% CI 0.93-1.03) and there was no relationship between stent type and death among patients with CLI (HR 0.97; 95% CI 0.92-1.03) or acute limb ischemia (HR 0.99; 95% CI 0.81-1.21).

If I’m treating a claudicator, I may be very careful about what I use and how much I use, but if I'm treating a CLI patient, they need the best I've got and right now that’s the paclitaxel devices. Christopher J. White

To TCTMD, Secemsky said the data support the continued use of paclitaxel-based devices, particularly in the CLI population.

“From a revascularization standpoint they have as much or more to gain than non-CLI patients,” he said. “On the flip side, the mortality in the CLI population is so high already that as long as we're not expediting their mortality with a device, it's probably reasonable to give them the best patency possible.”

White said the Medicare data are probably most relevant for clinicians who are being risk-averse right now in the wake of the safety concerns that have been raised.

“It's very important that we message the endovascular community that this is not a problem for CLI,” he said. “If I’m treating a claudicator, I may be very careful about what I use and how much I use, but if I'm treating a CLI patient, they need the best I've got and right now that’s the paclitaxel devices.”

Secemsky agreed, noting that there is growing skepticism about the mortality concerns and how they will ultimately impact practice once all the investigations have been completed, including by the FDA, which has already weighed in with a letter to clinicians.

“I think if we were going to stop using [paclitaxel devices] in a population it would make sense to do that in claudicants with a strong prognosis and [in whom] we wouldn't want to complicate that,” he observed.

White said while changes to practice could be in the offing if further research confirms the troubling meta-analysis, the mortality signal continues to evade understanding for the time being.

"We may not find the exact mechanism behind this for a very long time, but if people getting paclitaxel are dying at a higher rate than those who aren't then we need to rethink how we use paclitaxel," he said. "We may see a real push toward a lower-dose balloon in terms of what we recommend for people going forward."