番茄社区

German health insurance data support the long-term safety of paclitaxel-based drug-eluting stents (DES) and drug-coated balloons (DCBs) as used in peripheral arteries, researchers reported.

Paclitaxel stents were not significantly associated with death by 7 years (adjusted HR 1.10, 95% CI 0.84-1.44) or even 8-11 years (adjusted HR 0.64, 95% CI 0.40-1.01) compared with non-drug-eluting devices, according to a group led by Eva Freisinger, MD, of University Hospital Muenster, Germany.

Among balloons, paclitaxel devices were even associated with lower long-term mortality risk over the first year (adjusted HR 0.92, 95% CI 0.89-0.94), followed by no clear difference at 7 years (adjusted HR 0.88, 95% CI 0.70-1.12) or at 8-11 years (adjusted HR 1.02, 95% CI 0.74-1.40) in the study published online in .

Drug-eluting devices (DED) are "safe for endovascular therapy of the lower limbs," Freisinger's team concluded from their "real-world safety analysis" that covered the period from the 2007 introduction of these devices for peripheral artery disease (PAD) to the end of 2017.

Overall mortality reached 41.9% over the study period -- median follow-up being more than 7.5 years -- which the authors attributed to the 42.7% rate of chronic limb-threatening ischemia in the cohort from the start.

"Overall, the analysis is reassuring that in the real world there does not seem to be the same signal of risk as seen in the Katsanos meta-analysis," commented Aloke Finn, MD, of the CVPath Institute and University of Maryland, Baltimore, adding that the new data are in line with an analysis of the U.S. Medicare database published earlier this year.

This is the latest study to contradict suggesting a nearly two-fold increase in long-term mortality risk after paclitaxel device implantation for PAD. That meta-analysis has since been criticized for not including patient-level data and relying on studies with substantial loss to follow-up.

In contrast, the present study "contains no missing values," Freisinger and colleagues said. And because the exact coding, for example of DES versus bare metal stent (BMS), triggers a marked increase in reimbursement in the German system, "completeness of the applicable codes could be expected to be very high."

Konstantinos Katsanos, MD, of Patras University Hospital in Rio, Greece -- lead author of last year's meta-analysis -- called the German study an "elegant statistical analysis" but highlighted the difference in case-mix compared to his team's and the FDA's datasets.

This points to the "inconsistency between evidence from randomized trials (high internal validity and little selection bias during treatment allocation) compared to real-world administrative data (unselected patients with most likely a lot of selection bias during treatment choice)," he told 番茄社区.

"I think that the results appear reassuring but I am not sure whether they settle the question," Katsanos said.

The FDA is conducting its own safety investigation of these devices and, as of June 2019, determined that the mortality signal appears real.

Finn pointed out that no clear mechanism accounts for increased mortality, suggesting these devices "should continue to be part of the treatment of patients with symptomatic PAD."

Included in Freisinger's study were 64,771 people who were covered by Germany's Barmer Health Insurance (average age 72 years; over 45.3% women) and had a first endovascular revascularization from 2007 to 2015. The vast majority had undergone revascularization with BMS or "plain old balloon angioplasty" (POBA).

DES, DCB, BMS, and POBA groups had distinct baseline characteristics, requiring multivariable adjustment for comparison.

The German billing codes did not provide reasons for endovascular treatment, the investigators acknowledged. Nor could they retrieve details on the device manufacturer, paclitaxel load, or length of each device.

"Beyond the eighth year past DED application, the model faces the limitations of decreasing sample sizes for DES and DCB due to the timeliness of the index period," they added.

Comment