US Food and Drug Administration Perspective on “Mortality and Paclitaxel-Coated Devices: An Individual Patient Data Meta-Analysis”

atsanos et al 1 identified a late mortality signal associated with paclitaxelcontaining devices that was similarly observed in a subsequent US Food and Drug Administration (FDA) review of US-approved devices 2,3 and discussed by multiple stakeholders at an advisory meeting in 2019. 4Despite these analyses, uncertainty remains regarding the magnitude of the risk and the mechanism for mortality.The FDA communicated that clinicians should consider the increased rate of long-term mortality when making treatment recommendations 4,5 and implemented updated labeling for this device class to communicate the risk.The uncertainty regarding the benefit-risk profile of these devices continues to affect patients with peripheral arterial disease and their physicians.The article by Rocha-Singh et al 6 on behalf of the Vascular Interventional Advances (VIVA) Physicians in this issue of Circulation provides an independent mortality assessment of manufacturer-supplied, deidentified patient-level data from randomized, controlled trials (RCTs) in an effort to provide additional clarity regarding the magnitude and mechanism for the identified mortality risk.The analysis includes data from 2185 patients in 8 RCTs that evaluated FDAapproved paclitaxel-coated balloons and stents when used to treat peripheral arterial disease.The primary analysis reported a 38% relative increase in mortality hazard (hazard ratio, 1.38 [95% CI, 1.06-1.80])through 5 years for paclitaxelcoated devices compared with uncoated control devices.The rate decreased to 27% (hazard ratio, 1.27 [95% CI, 1.03-1.58])after the recovery of additional vital status information.These data suggest increased mortality similar to the reports by Katsanos et al 1 and the FDA (relative risk of 1.93 and 1.57, respectively).No mechanism of action for the late-term mortality was identified.The VIVA analysis also included an assessment for dose response related to mortality (in tertile dose groupings), examined using a proportional hazards model with an adjustment for covariates using propensity scores.A clear dose-response association could not be determined.The results of the VIVA analysis provide important clarifying information on the late-term mortality signal that are consistent with the conclusions of Katsanos et al 1 and the FDA regarding the increased risk of mortality for patients treated with paclitaxel-containing devices.Their collaborative effort included the incorporation of patient-level, timeto-event data and baseline covariates across studies (8 RCTs) and the use of statistical models to account for both study-level and patient-level variation by considering study-level random effects and patient-level covariates.Although the primary analysis set was the intent-to-treat population, the article also summarizes an as-treated population analysis to account for bias attributable to