No Long-Term Increased Mortality Seen in Medicare PAD Patients Treated With Peripheral DES
March 14, 2019—Findings from an investigation examining long-term mortality between drug-eluting stent (DES) and bare-metal stent (BMS) implantation in patients with a principal diagnosis of peripheral artery disease (PAD) were published online ahead of print by Eric A. Secemsky, MD, et al as a letter in the Journal of the American College of Cardiology (JACC).
The study was composed of Medicare patients with a diagnosis of PAD as identified by International Classification of Diseases–Ninth Revision–Clinical Modification claims codes. The study included procedures performed between December 1, 2012—corresponding to the approval of the first peripheral DES (Zilver PTX, Cook Medical)—and September 30, 2015. These procedures were identified using International Classification of Diseases–Ninth Revision–Procedure Coding System (ICD-9-PCS) codes.
Dr. Secemsky and colleagues found no evidence of increased long-term mortality following peripheral artery revascularization with DES compared with BMS at a median 2 years of follow-up, with the longest follow-up extending to 4.1 years. They stated that this finding suggests the safety of these devices in routine practice. Additionally, the lack of an association between peripheral DES implantation and mortality remained after multivariable adjustment and among patients with critical limb ischemia (CLI) or acute limb ischemia (ALI).
As stated in JACC, this study was motivated by the meta-analysis published by Konstantinos Katsanos, MD, et al online in December 2018 in the Journal of the American Heart Association that found an association between peripheral paclitaxel-coated devices and increased long-term mortality, with an 87% increased risk of all-cause death at 5 years. However, these findings have not been replicated in other data sources with extended follow-up, noted Dr. Secemsky and colleagues.
In Dr. Secemsky's study, the cumulative incidence of death through December 31, 2016, was calculated using Kaplan-Meier methods, and log-rank tests were used to evaluate for differences between groups. Multivariable Cox regression was used to calculate hazard ratios (HR), adjusted for age, sex, and comorbidities. Subgroup analyses were performed among patients with CLI and ALI. All P values were considered significant (P < .05).
As reported by the investigators in JACC, the study was composed of 51,456 patients who underwent peripheral stenting (average age, 72.8 ± 10.5 years; 54.2% were men; 59.7% had CLI; 7.1% had ALI; and 39.3% were diabetic). During the study, there was a gradual increase in the use of peripheral DES.
The investigators found the following:
- Through 4.1 years, patients treated with DES versus BMS had similar mortality (51.7% vs 50.1%; log-rank P value = .16). This relationship persisted after stratification by CLI.
- There was no association between stent type and mortality after multivariable adjustment (HR for DES vs BMS, 0.98; 95% confidence interval [CI], 0.93–1.03; P = .53).
- There was no adjusted relationship between stent type and death among patients with CLI (HR, 0.97; 95% CI, 0.92–1.03; P = .32) or ALI (HR, 0.99; 95% CI, 0.81–1.21; P = .95).
Dr. Secemsky and colleagues advised that limitations of their analysis include the inability to localize the lesion of interest; possible misclassification with use of claims codes; lack of data on outpatient procedures and drug-coated balloons (no ICD-9-PCS code); the potential influence of unmeasured confounding; and the inability to determine specific causes of death. In addition, the CMS population was older and had more comorbidities, including CLI, compared with the Katsanos et al meta-analysis, advised the investigators in JACC.
In an interview with Endovascular Today, Dr. Secemsky advised that a distinction between this study and a prospective randomized trial is the adjudication of the determination of death. Dr. Secemsky commented, "The main difference in our analysis is that Medicare has a vital status record that captures deaths for all beneficiaries. So, any patient who dies in follow-up will no longer be processing claims in the Medicare databases and will be recorded in the vital status files. As such, we can confidently follow the majority of patients we studied within the Medicare database for the duration of follow-up we have available.
"On the flip side, we are unable to determine the specific cause of death from these files, only that they were deceased. In randomized controlled trials, mortality data can be obtained in many ways, depending on how the study was set up. Typically, this will occur while arranging part of preplanned follow-up interviews or visits, or death notices received by the recruiting centers. However, if a patient has completed a trial, let's say at 1 year, then the study team might not be able to determine or obtain death data 2 or 5 years later. As such, follow-up may be variable and as such, mortality data might only be available for select patients who remained in contact longitudinally."
For links to all of Endovascular Today's coverage and other key documents on the continuing discussion regarding paclitaxel in PAD, please visit this page.