Balloon angioplasty with or without stent placement is a common revascularization strategy in patients with femoropopliteal lesions and associated claudication that is refractory to optimal medical therapy. Yet restenosis resulting in symptom recurrence is common and often requires reintervention.

Paclitaxel is an antiproliferative agent applied to the surface of an angioplasty balloon or stent that has been shown to reduce the risk of vessel restenosis and associated reinterventions compared with uncoated devices (UCDs), without increased safety risks, in patients with femoropopliteal atherosclerotic lesions.^1,2 Consequently, paclitaxel-coated devices (PCDs) have become widely adopted by vascular specialists for treatment of peripheral artery disease.

Meta-analysis Raises Important Safety Questions

In December 2018, a meta-analysis of randomized controlled trials (RCTs) published by Katsanos and colleagues reported that PCDs increased the risk of death at two to five years follow-up in patients with femoropopliteal disease.³ At the five-year follow-up period, the relative risk of mortality in three studies with PCDs was nearly double that of UCDs. This conclusion was novel and highly controversial.

The publication of this meta-analysis prompted widespread debate in the vascular community and had immediate impact on patient practice. The BASIL-3 and SWEDEPAD randomized trials of PCDs vs. UCDs temporarily halted patient enrollment within days of publication of the meta-analysis.

Interestingly, the decision to suspend enrollment in the SWEDEPAD trial was based, in part, on the meta-analysis findings but also on results of an interim safety analysis in the trial involving 1,480 patients with intermittent claudication and 810 with critical limb ischemia.⁴ These interim safety results have not been published, but it seems possible that an adverse safety signal may have been detected.

In January 2019, the U.S. Food and Drug Administration (FDA) issued a letter to health care providers in which they advised doctors to continue monitoring patients who were treated with PCDs and they stated that the benefits of these devices continued to outweigh the risks when used as indicated.⁵

The FDA is also reviewing long-term, patient-level mortality data made available from manufacturers of these devices to more rigorously explore the possible link between paclitaxel and excess mortality. Until the results of these analyses become available, the methods and findings of the meta-analysis in question should be critically evaluated in a balanced manner.

Critical Review of the Katsanos Meta-analysis

The several strengths of the Katsanos meta-analysis were inclusion of RCTs with at least one year of follow-up data in which PCDs were compared with UCDs; extraction and analysis of mortality data at one year, two years, and four to five years; and performance of various subgroup and sensitivity analyses that corroborated the main conclusions of the review.

Inspection of the crude event rates among the trials included in the meta-analysis reveals a consistent pattern for more deaths in the paclitaxel group. An advantage of a meta-analysis in general is the ability to increase sample size by pooling data from multiple studies that may allow detection of risks which could be missed in individual studies due to a lack of statistical power. In that respect, the meta-analysis proved useful because the identified mortality signal could only have been detected in a clinical trial that enrolled thousands of patients.

However, the meta-analysis also suffered from several important limitations. Intention-to-treat methods were not used which may have biased the results. Crude mortality rates calculated by the number of deaths divided by the number of patients enrolled were reported. Since loss to follow-up is significant in peripheral vascular disease trials, especially with increasing follow-up, failure to account for attrition using survival analyses is problematic. Well-known methods exist to estimate survival outcomes for use in a meta-analysis even when only crude mortality rates are provided in the original articles.⁶

A causal mechanism by which paclitaxel may have conferred excess mortality was not established in the present meta-analysis. Since the paclitaxel half-life when applied to peripheral devices ranges from days to months, it remains unclear how the drug could contribute to excess mortality at two years and beyond.

Finally, a meta-analysis of aggregate data does not allow for many types of detailed analyses that patient-level data provide. Interestingly, similar controversies have occurred with paclitaxel-eluting stents for coronary interventions. A meta-analysis of RCTs concluded that paclitaxel-eluting stents for coronary interventions increased mortality risk relative to bare metal stents.⁷

Yet upon re-analysis using an individual patient data (IPD) meta-analysis, mortality risk was comparable between treatment groups.⁸ These results highlight the role of an aggregate meta-analysis as a tool for hypothesis generation only, in which conclusions must later be confirmed using more robust methods that utilize IPD.

Recent Study Results

Since the publication of the Katsanos meta-analysis, results of several relevant studies have become available that add to the current evidence base related to paclitaxel. Schneider et al.,⁹ reported results of a propensity-score weighted IPD meta-analysis in which mortality rates through five years were 9.3 percent with drug-coated balloons and 11.2 percent with percutaneous transluminal angioplasty (PTA) (p=0.40) among patients with femoropopliteal disease.

Secemsky et al.,¹⁰ reported results of PCD vs. UCD among Medicare beneficiaries undergoing femoropopliteal revascularization in which survival was slightly lower with PCD (32.5 vs. 34.3 percent; p=0.007) over 1.6 years follow-up.

Additionally, numerous recent presentations at the LINC and ISET annual meetings have reported no increased mortality risk of PCD in patient-level meta-analyses. However, these recent reports typically included only the outcomes with specific devices and some reports had limited follow-up. Ultimately, these recent reports in and of themselves do not provide compelling evidence that PCDs are entirely benign over the long term.

In fact, corrections to previously published data have been recently issued for two PCDs that add to the controversy. Medtronic issued a press release stating that a data programming error was made in the IN.PACT Global postmarket study, where an unspecified number of patient deaths were inadvertently omitted from the analysis.¹¹

The company stated that, after correcting the error, five-year mortality rates in the study remained comparable between the DCB and PTA groups. While the magnitude of this error on previously reported mortality rates is currently unclear, two prior publications and an IPD meta-analysis are impacted. Data corrections to these reports are planned for public dissemination soon.

Another correction to previously published data was made regarding the five-year mortality rates in the Zilver PTX trial comparing DES to PTA. The original paper, published in 2016, reported five-year mortality rates of 10.2 percent with DES and 16.9 percent with PTA (p=0.03).¹²

The recent correction states that these rates were reversed in the paper and that the correct rates are 16.9 percent with DES and 10.2 percent with PTA.¹³ This new conclusion is more intuitive given a prior publication reporting a two-fold increase in two-year mortality rates (7.6 percent vs. 3.4 percent) with DES from the same trial.¹⁴

Overall, due to the mixed results from these numerous recent reports that have become available since the Katsanos meta-analysis was released, it appears that a definitive conclusion regarding PCD safety remains elusive.

Where Do We Go From Here?

While efforts to provide pooled IPD analyses from all industry stakeholders continue, trials not sponsored by industry should also be considered for inclusion in these analyses. Attempts should be made to acquire IPD data related to baseline characteristics, procedure and device details, and mortality outcomes from all these studies to facilitate rigorous and comprehensive reporting. Such an analysis is particularly important to dissuade concern that results may be influenced by industry bias or publication bias.

Of the ongoing or completed RCTs of PCD for femoropopliteal disease, only a fraction of available data on PCD was likely available for use in the aggregate meta-analysis of Katsanos and colleagues.

Ultimately, given the existence of approximately 35 relevant RCTs that have explored PCD vs. UCD, a monumental amount of unpublished data is likely available that, once analyzed and published, would provide more definitive guidance regarding the benefits and risks associated with PCD for peripheral applications.

Table 1: Ongoing or Completed Trials of PCDs For Femoropopliteal Disease
Study Name	Clinical Trial Registration	Included in Katsanos meta-analysis? *
ACOART I	NCT01850056	Yes
BASIL-3 †	ISRCTN14469736	No
BATTLE	NCT02004951	Yes
BIOLUX P-I	NCT01221610	Yes
CONSEQUENT	NCT01970579	Yes
DEBATE-SFA	NCT01556542	Yes
DEBATE-IN-SFA	UMIN000010071	Yes
DEBELLUM	Not registered	Yes
DRECOREST	NCT03023098	Yes
EFFPAC	NCT02540018	Yes
FAIR	NCT01305070	Yes
FEMPAC	NCT00472472	Yes
FREEWAY	NCT01960647	Yes
ILLUMENATE	NCT01858428; NCT03421561	Yes
ILLUMENATE EU	NCT01858363	Yes
INVADER MRI	NCT02807779	No
IN.PACT SFA I & II	NCT01175850; NCT01566461	Yes
IN.PACT SFA JAPAN	NCT01947478	Yes
ISAR-PEBIS	NCT01083394	Yes
ISAR-STATH	NCT00986752	Yes
LEGFLOW-2015-01	NCT02965677	No
LEVANT I	NCT00930813	Yes
LEVANT II	NCT01412541	Yes
LUTONIX JAPAN	UMIN000014673	Yes
MAGNIFICENT	NCT02710656	No
MDT-2113 SFA JAPAN	NCT01947478	No
PACIFIER	NCT01083030	Yes
PACUBA	NCT01247402	Yes
RANGER SFA	NCT02013193	Yes
RANGER II SFA	NCT03064126	No
RAPID	ISRCTN47846578	Yes
SINGA-PACLI	NCT02129634	No
SWEDEPAD †‡	CT02051088	No
THUNDER	NCT00156624	Yes
ZILVER PTX	NCT00120406	Yes

*The Katsanos meta-analysis included the FINN-PTX trial, which is not a relevant study because it compared paclitaxel-eluting stent vs. bypass surgery using expanded polytetrafluoroethylene grafts.
†Patient enrollment temporarily suspended.
‡Eligibility criteria allows above-knee and/or below-knee lesions.

---

This article was authored by Jihad A. Mustapha, MD, FACC; Fadi A. Saab, MD, FACC; and Larry E. Miller, PhD. Mustapha and Saab are interventional cardiologists specializing in endovascular revascularization of peripheral artery disease, particularly critical limb ischemia, at Advanced Cardiac and Vascular Amputation Prevention Centers, Grand Rapids, MI. Each also serves as clinical associate professor of medicine at Michigan State University College of Osteopathic Medicine. Miller is president of Miller Scientific Consulting, Inc., Asheville, NC.

---

References

1. Jaff MR, Nelson T, Ferko N, et al. J Vasc Interv Radiol 2017;28:1617-27.
2. Katsanos K, Spiliopoulos S, Karunanithy N, et al. J Vasc Surg 2014;59:1123-33.
3. Katsanos K, Spiliopoulos S, Kitrou P, et al. J Am Heart Assoc 2018;7:e011245.
4. Temporary Halt of recruitment to the SWEDEPAD trials. Available here. Accessed Feb 14, 2019.
5. U.S. FDA. Treatment of peripheral arterial disease with Paclitaxel-coated balloons and paclitaxel-eluting stents potentially associated with increased mortality - letter to health care providers. Available here. Accessed Feb 19, 2019.
6. Tierney JF, Stewart LA, Ghersi D, et al. Trials 2007;8:16.
7. Nordmann AJ, Briel M, Bucher HC. Eur Heart J 2006;27:2784-814.
8. Stone GW, Moses JW, Ellis SG, et al. N Engl J Med 2007;356:998-1008.
9. Schneider PA, Laird JR, Doros G, et al. J Am Coll Cardiol 2019;Jan 25:[Epub ahead of print].
10. Secemsky EA, Kundi H, Weinberg I, et al. JAMA Cardiol 2019;Feb 12:[E-pub ahead of print].
11. Medtronic. Medtronic statement regarding revised IN.PACT post market study data. Available here. Accessed Feb 21, 2019.
12. Dake MD, Ansel GM, Jaff MR, et al. Circulation 2016;133:1472-83.
13. [No authors listed]. Correction. Circulation 2019;139:e42.
14. Dake MD, Ansel GM, Jaff MR, et al. J Am Coll Cardiol 2013;61:2417-27.

Keywords: ACC Publications, Cardiology Magazine, Angioplasty, Angioplasty, Balloon, Angioplasty, Balloon, Coronary, Arteries, Comorbidity, Drug-Eluting Stents, Half-Life, Follow-Up Studies, Intention to Treat Analysis, Intermittent Claudication, Medicare, Metals, Ocimum basilicum, Paclitaxel, Peripheral Arterial Disease, Peripheral Vascular Diseases, Physicians, Publication Bias, Propensity Score, Registries, Reproducibility of Results, Risk Assessment, Risk, Sample Size, Stents, Survival Analysis, United States Food and Drug Administration, Specialization

< Back to Listings