No Evidence of Paclitaxel-Related Increased Mortality Found in Medicare Database Analysis at 600 Days
February 14, 2019—A study published online in JAMA Cardiology by Eric A. Secemsky, MD, et al showed no evidence of increased all-cause mortality following femoropopliteal artery revascularization using paclitaxel drug-eluting stents (DESs) or drug-coated balloons (DCBs) versus non–drug-coated devices. The study seeks to explore concerns raised by the meta-analysis by Katsanos et al regarding the potential for increased mortality associated with the use of paclitaxel in patients with peripheral artery disease.
The investigators evaluated data from 16,560 patients in the US Medicare database (mean age, 72.9 years; 46.7% men) treated for femoropopliteal artery revascularization at 1,883 hospitals. Drug-coated devices were used in 5,989 (36.2%) of cases and were shown to be associated with a lower cumulative incidence of all-cause mortality compared with treatment using non–drug-coated devices through 600 days (32.5% vs 34.3%, respectively; log-rank P = .007).
As reported in JAMA Cardiology, a multivariable adjustment was conducted and showed no association between drug-coated devices and all-cause mortality (adjusted hazard ratio, 0.97; 95% confidence interval, 0.91–1.04; P = .43).
“These are critically important data that need to help provide reassurance of the safety of these devices,” said Dr. Secemksy, in comments to Endovascular Today. Dr. Secemksy is Director of Vascular Intervention at Beth Israel Deaconess Medical Center, Harvard Medical School, in Boston, Massachusetts.
This research comes in response to the recent meta-analysis of randomized controlled trial (RCT) data from Katsanos et al that had shown an increased long-term mortality in patients treated with paclitaxel-coated devices. The data published by Secemsky et al are significant because they are the first data set published on this topic evaluating real-world data of patients treated with paclitaxel-coated devices, rather than from registry and RCT data.
“The majority of real-world PAD patients are older, sicker, and with multiple comorbid conditions,” said Dr. Secemsky. “These are also the patients with more complex disease and recurrent need for intervention, and thus are the most likely to benefit from drug-coated technology.”
In a subgroup analysis, patients with critical limb ischemia (n = 8,450) who were treated with DCBs or DESs, had a lower cumulative incidence of mortality through 600 days postprocedure compared with CLI patients treated with non–drug-coated devices (38.1% vs 40.1%, respectively; log-rank P = .04). Among patients without CLI, there was no statistical difference observed in cumulative incidence of survival between those treated with drug-coated devices and non–drug coated devices (26.5% vs 29%, respectively; log-rank P = .07).
"Oftentimes, trials are restrictive in regard to enrollment and rarely generalizable to clinical practice. As such, demonstrating that drug-coated devices did not reduce survival in a more representative patient population is of great importance and should be reassuring to both clinicians and patients,” commented Dr. Secemsky.
However, while the study period was 600 days, in the current paper, Secemsky et al were only able to provide mortality data on patients with a mean follow-up of 389 days (interquartile range, 277–508 days). The Katsanos et al meta-analysis also found no difference in mortality at 1 year, but did observe higher mortality at 2 years, and higher still at a longer 4- and 5-year endpoint.
In an accompanying invited commentary published online in JAMA Cardiology, Jay Giri, MD, MPH, emphasized the inclusion of robust data presented in Secemsky et al, in particular that all subjects evaluated were exclusively inpatient with the majority having a high burden of CLI. This was also in contradistinction to the studies evaluated by the Katsanos et al meta-analysis, which included only 11% CLI patients. However, Dr. Giri, Assistant Professor of Medicine at the Hospital of the University of Pennsylvania, in Philadelphia, Pennsylvania, also advised that the median follow-up length may not have been sufficient enough to detect late paclitaxel toxicity.
“Ideally, our study would have been capable of providing similar, if not longer, follow-up after revascularization with drug-coated devices as was reported in the Katsanos et al study,” commented Dr. Secemsky to Endovascular Today. He noted that the current ICD (International Classification of Diseases)-10 codes were not implemented by the Centers for Medicare & Medicaid Services until late 2015, which factored into the limitation of the study’s follow-up period. “However, our study is both strengthened and limited by the use of ICD-10 claims codes. Drug-coated balloons have a specific ICD-10 claims code, which is not the case for ICD-9 codes. The use of these codes allowed us to perform this study with the confidence that we were studying the correct device.”
Additionally, Dr. Secemsky noted that the Katsanos el al meta-analysis reported long-term mortality data at three discrete timepoints: 1, 2, and 4-5 years, with the caution that its source studies were not designed or powered to evaluate mortality this far out. “These findings are really hypothesis-generating at best,” said Dr. Secemsky.
“The 2-year findings included 12 RCTs and more than 2,300 patients, and this is where the authors found a 3.5% mortality difference not seen at year 1,” he continued. “Although our study falls short of this 2-year mark, we do shed light on whether there is a signal of mortality occurring between 1 and 2 years. This is important as, if this relationship is real, we would expect to see a gradual separation in survival curves during this time period.
“In our analysis, although the median follow-up was 389 days, we still analyzed more than 3,000 patients at 540 days and nearly 1,000 patients through 600 days. During this time, we saw no signal that survival was worsening among patients treated with drug-coated devices,” said Dr. Secemsky. “We would be surprised that if we were able to add another 130 days of follow-up, we would see anything close to a 3.5% mortality difference.”
The study authors will have additional follow-up data published later this month, as well as plans to continue to update the current data they published as longer follow-up becomes available.
“The FDA message was spot-on—it’s important to weigh both the risks and benefits with any treatment or intervention,” said Dr. Secemksy when asked how he would advise practicing clinicians who are trying to balance these emerging data. “Until we have seen enough data to fully exonerate this relationship, we should continue to counsel patients about the concerns raised with drug-coated devices.
For more on current data and key articles on paclitaxel delivery devices, please visit our continuing coverage page here.