No Long-term Mortality Risk for Paclitaxel-Based Peripheral Devices: Medicare Advantage Cohort

There is no evidence that paclitaxel-coated devices used for the treatment of femoropopliteal artery disease increases the long-term risk of death when compared with an uncoated balloon or stent, according to a large analysis of Medicare Advantage patients. The findings were presented in a late-breaking trial livestream hosted June 25, 2020, by VIVA Physicians.

“I think this is a really important piece in the paclitaxel story because this is a data set that the FDA found value in and wanted to have available to help make decisions about moving forward with these devices,” presenter Eric Secemsky, MD, MSc (Beth Israel Deaconess Medical Center, Boston, MA), told TCTMD.

The new analysis followed a prespecified analytic plan requested by the US Food and Drug Administration and is one of several similar Medicare cohort studies that have been conducted over the last 16 months. All of these efforts were aimed at investigating a mortality signal first raised by Konstantinos Katsanos, MD, PhD (Patras University Hospital, Rion, Greece), in a 2018 meta-analysis. Katsanos et al showed there was a 6.6% increased absolute risk of death in paclitaxel-treated patients with PAD versus those treated with an uncoated device over 4 to 5 years. The difference in mortality began to emerge around 2 years after the intervention.

When Secemsky presented early Medicare data at an FDA advisory panel meeting in June 2019, it included 152,473 fee-for-service Medicare beneficiaries and had a median follow-up of 799 days, with the longest being 1,573 days. It did not include patients covered under Medicare Advantage plans, of which there are several that represent over one-third of all Medicare enrollees. The FDA therefore requested an analysis looking at all-cause death in those patients.

Still No Signal

For the new study, Secemsky and colleagues examined data on 16,796 Optum Medicare Advantage patients (average age 73.3 years; 44.4% women) who were treated with paclitaxel-coated peripheral devices or uncoated devices from April 2015 through December 2017. The median follow-up was 2.66 years, with the longest being 4.75 years.

In his presentation, Secemsky noted that Medicare Advantage patients have higher self-reported health status and lower cognitive and functional impairments compared with the traditional Medicare group, and they also tended to be about 3 years younger.

At baseline, patients in the paclitaxel and uncoated device groups were similar with the exception of a greater use of atherectomy and P2Y12 inhibitors in the paclitaxel group. However, inverse probability of treatment weighting balanced out the groups by weighting patient-level data according to observed characteristics. This included a “look back” in which comorbidities and cardiovascular medications were accounted for in the 3 months prior to treatment.

The cumulative incidence of all-cause mortality (primary endpoint) was 45.8% in the drug-coated group and 43.2% in the uncoated group (adjusted HR 1.03; 95% CI 0.96-1.10). The results remained the same when they were further parsed to account for treatment with paclitaxel balloons versus percutaneous transluminal angioplasty (adjusted HR 1.00; 95% CI 0.92-1.08) or paclitaxel stents versus PTA (adjusted HR 1.02; 95% CI 0.88-1.18). Additionally, no differences were seen in prespecified subgroup analyses accounting for treatment setting (outpatient vs inpatient), PAD severity, age, sex, or comorbidity burden (interaction P > 0.05 for all).

Are We There Yet?

Sean P. Lyden, MD (Cleveland Clinic, OH), who moderated the livestream, noted that concerns have been raised since the first of the Medicare analyses about whether the population is a suitable comparison to the randomized trial data in which the paclitaxel signal was initially seen.

“I think it’s amazing that you’re now showing with similar [or] lesser comorbidities and no CLI that there’s no difference [in mortality],” he said.

Given the circus-like atmosphere that has surrounded attempts to understand the paclitaxel mortality signal that emerged from the original meta-analysis, panelist Misti Malone, PhD (US Food and Drug Administration, Silver Spring, MD), asked Secemsky if his research has given him insight into what the approach should be if another signal related to a drug or device ever were to appear again.

He noted that while researchers have been tested by paclitaxel, they have learned enough to be ready to respond if a similar situation arises.

“We've learned how to not only interpret claims data and observational data in a more sophisticated fashion, but really to create a template about . . . how to analyze the data and stress the data sets, stress the analysis, to make sure that what we're finding is of the highest confidence,” he reiterated to TCTMD.

Secemsky added that the analyses of the various paclitaxel trial data sets and the Medicare population suggests that it is time to think about “moving forward again with treating our patients with devices that really can benefit them.” In its last letter to healthcare providers, issued in August 2019, the FDA continued to express concern about the mortality signal and stressed the importance of discussing it with patients and reserving use of the devices for those at particularly high risk for restenosis and repeat femoropopliteal interventions.

“I think it's time now, a year and a half into this, to really reevaluate where we stand, reevaluate our position and the FDA's position and societies’ position on these devices,” he noted. “Everything that was asked in June 2019 has really been respected and I think that we have a lot of new and important data that has not shown that signal of harm.”

Evaluation of Medicare patients with PAD who have received paclitaxel is continuing in the SAFE-PAD study, which is being conducted in conjunction with the FDA and will follow the fee-for-service cohort for a median of 5 years or longer. It will include sensitivity analyses for unmeasured confounding and further subgroup analyses. New data from that study are expected soon.