Mortality After Paclitaxel-Coated Device Use in Patients With Chronic Limb-Threatening Ischemia: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Abstract

Purpose:

To report the risk of all-cause mortality after treatment with paclitaxel-coated devices vs uncoated controls in patients with chronic limb-threatening ischemia (CLTI).

Materials and Methods:

A search of the PubMed, Embase, Cochrane, CINAHL, DARE, and PROSPERO databases was made on 5 November 2019 to identify randomized controlled trials (RCT) using intention-to-treat analysis to compare a paclitaxel-coated device to an uncoated device in PAD patients having clinical follow-up of at least 6 months. Half of the study population had to have CLTI or extractable data on the CLTI subgroup if <50%. The search identified 11 trials having 1450 patients randomized to a paclitaxel-coated device (n=866) or an uncoated control (n=584). There were 1367 (94.3%) patients with CLTI (range 10–429). The single endpoint was all-cause mortality, which was analyzed by pooling the mortality data in a DerSimonian and Laird random effects model. Summary statistics are expressed as relative risk ratios (RR) with a 95% confidence interval (CI).

Results:

The mean follow-up was 25.6 months (range 6–60); 10 of 11 studies reported a minimum 12-month follow-up. There were 161 (18.6%) deaths among 866 subjects in the paclitaxel device group and 116 deaths among 584 (19.9%) subjects in the non-coated control group (RR 0.93, 95% CI 0.78 to 1.12, p=0.45).

Conclusion:

There was no observed difference in short- to midterm mortality among a pooled patient population of predominately CLTI patients treated with paclitaxel-coated balloons or stents compared with uncoated controls.

Keywords

Balloon angioplasty chronic limb-threatening ischemia critical limb ischemia drug-coated balloon drug-eluting balloon drug-releasing balloon paclitaxel drug-eluting stent paclitaxel-coated balloon peripheral artery disease percutaneous transluminal angioplasty

Introduction

Peripheral artery disease (PAD) is a global epidemic that affects more than 230 million people.¹ It may be asymptomatic or associated with symptoms that range from intermittent claudication to chronic limb-threatening ischemia (CLTI), a condition characterized by rest pain and ulceration.² In contrast to claudication, CLTI is associated with significant risk to life and limb.²

Endovascular revascularization is often used as the primary treatment method for CLTI.³ Its goal is to revascularize the ischemic limb in order to facilitate wound healing, resolve pain, reduce the need for amputation, and improve quality of life. It has also been found to be cost effective.^4-6 However, it is common for the target artery to develop restenosis, which may lead to recurrent symptoms.³ The antiproliferative properties of paclitaxel-coated balloons and stents are an effective means to increase the durability of the treatment by reducing intimal hyperplasia and negative remodeling.^7,8 Devices coated with paclitaxel are known to effectively reduce the need for reintervention.^7-9

A recent meta-analysis reported an increased mortality risk at 2 and 5 years after the use of paclitaxel-coated devices in the femoropopliteal segment of patients suffering predominantly from claudication.¹⁰ Although no plausible biological mechanism for paclitaxel toxicity has been proposed and no dose-related biologic gradient demonstrated, the potential for increased late mortality is a safety concern that demands research into other patient populations in whom these devices are used. The aim of this study was to determine the rate of all-cause mortality in a CLTI population of patients treated with paclitaxel-coated vs uncoated devices.

Materials and Methods

Literature Search

The PubMed, Embase, Cochrane, CINAHL, DARE, and PROSPERO databases were searched on 5 November 2019 to identify randomized controlled trials (RCT) investigating the use of drug-coated balloons or stents for the treatment of patients with CLTI. The search was conducted without restriction on time frame, publication status, or language. The search terms used included “critical limb ischemia,” “peripheral artery disease,” “peripheral vascular disease,” “arterial occlusive disease,” “ischemia,” “intermittent claudication,” “amputation,” “drug-eluting*,” “drug-coated*,” “drug elut*,” “paclitaxel,” “angioplasty,” “balloon*,” “bare-metal,” “metal,” “percutaneous,” “stents*,” “bare*,” as well as the corresponding Medical Subjects Headings (MeSH) with Boolean Syntax (Appendix 1). The reference list of each identified study was also screened for additional citations that might relate to the topic.

Studies were imported into a web-based software analysis package (Covidence systematic review software; Veritas Health Innovation, Melbourne, Australia). All were evaluated for relevance, quality, and risk of bias for this analysis. This systematic review protocol was registered with the PROSPERO database of systematic reviews (CRD42019141227; www.crd.york.ac.uk/PROSPERO/).

Inclusion and Exclusion Criteria

Studies were selected if they were an RCT using intention-to-treat analysis to compare a paclitaxel-coated device to an uncoated device in PAD patients having clinical follow-up of at least 6 months. Half of the study population had to have CLTI or extractable data on the CLTI subgroup if <50%. Excluded studies included coronary trials, nonrandomized studies, case reports, and duplicate studies.

Study Selection

The PRISMA search strategy¹¹ and trial selection are presented in Figure 1. The literature search yielded 1154 articles. After removing 304 duplicate studies, 850 abstracts were individually screened for relevance. A further 779 studies did not meet the predetermined inclusion criteria and were excluded, leaving 71 for full text screening. Eleven studies^9,12-20,23 were selected: 7 studies^9,12-17 in which the participants included a majority of CLTI patients (>50%) and 4 studies^18,19,20,23 having <50% CLTI subjects but with sufficient data for analysis.

Data Abstraction and Endpoints

Two authors (K.D. and M.L.G.) independently interrogated all included studies using a standardized data abstraction method. Details included the patient’s baseline characteristics, Rutherford category, antiplatelet regimen, lesion location, type of intervention, and mortality outcome. Results were verified by a third author (R.L.V.). Data were extracted from all manuscript content, including main text, tables, figures, and survival curves, then evaluated on an intention-to-treat basis. When there were multiple manuscripts from the same study, the latest publication with the longest follow-up period was used. If endpoint data were missing or conflicted with previous publications or long-term follow-up had been presented at a scientific congress but not published, the authors were contacted directly.

The primary outcome measurement was all-cause mortality, which was defined as any death after the procedure regardless of cause. The denominator used to calculate overall survival was the number of subjects randomized to each study arm (if ≥50% had CLTI) or the number of subjects with CLTI (if <50%). Sensitivity and subgroup analyses were performed to examine the interaction of various study variables with the mortality endpoint. No secondary endpoints were measured.

Risk of Bias

The risk of bias was assessed for each included study by 2 authors (K.D. and M.L.G.) independently using the revised Cochrane risk-of-bias tool for randomized trials,²¹ with a third author (R.L.V.) resolving any conflicts. This tool examines the studies over 7 methodological domains: random sequence number generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other bias (such as baseline imbalance or not reporting sample size calculations).

Statistical Analysis

Mortality rates were pooled in a DerSimonian and Laird random effects model²² to account for clinical and/or methodological heterogeneity. Summary statistics were expressed as relative risk ratios (RR) with a 95% confidence interval (CI). The chi-square test and the I² statistic were used to assess the level of statistical heterogeneity. An I² <25% was considered evidence of low statistical heterogeneity, 25% to 50% moderate, and >50% as high. Publication bias was evaluated using the inverted funnel plot method with visual estimation to determine asymmetry. Sensitivity and subgroup analyses were performed to examine mortality according to antiproliferative drug dose, the location and extent of the target lesion, as well as whether the paclitaxel-coated device was a balloon or stent. Statistical significance was set at p<0.05. If statistical significance was found, then the absolute risk difference and a number-needed-to-harm with associated 95% confidence intervals (CI) were calculated. Quantitative analysis of the included RCTs was performed using Review Manager (RevMan version 5.3.5.; The Cochrane Collaboration; http://community.cochrane.org/help/tools-and-software/revman-5 ), OpenMeta[Analyst] ( http://www.cebm.brown.edu/openmeta/ ), and SPSS (version 24; IBM Corporation, Armonk, NY, USA) software packages.

Results

Characteristics of Included Studies

Of the 11 studies, 8 were multicenter^9,15-20,23 and 3 were single-center^12-14 RCTs (Table 1). Five studies were single-blinded,^{9,16,17,19,20} 4 were non-blinded,^13-15,23 and 2^12,18 did not specify their blinding method. Nine studies^{12-15,17-20,23} employed a 1:1 randomization and 2 studies^9,16 employed a 2:1 randomization; all conducted an intention-to-treat analysis. A single study¹⁸ employed a 2-tier randomization system in which patients initially randomized to the control intervention who required a bailout stent were secondarily randomized between paclitaxel-coated and uncoated nitinol stents.

Table 1.

Characteristics of the Included Randomized Controlled Trials.

Author, Year	Study Design, Blinding	Na	CLTI, %	Relevant Inclusion Criteria	Relevant Exclusion Criteria	Arms	Drug Device	Follow-up, mo	Deaths, Paclitaxel / Control
Dake¹⁸ 2016	MC, NS	474	9	RC 2+, ≥50% stenosis, RVD 4–9 mm, lesion length ≤14 cm, ≥1 patent runoff vessel	Untreated >50% stenosis of the inflow tract, lesion pre-treatment with adjunctive devices, previous target vessel stenting	1:1 DES vs BA (BA failures randomized 1:1 DES vs BMS)	Zilver PTX	60	7 / 6
DeBoer¹⁹ 2017	MC, single-blinded	160	16	RC 2+, de novo SFA lesions ≥50 mm, occlusions or ≥50% stenosis on CTA or MRA or PSVR ≥2.4 on duplex, ≥1 patent BTK artery with uninterrupted flow to the pedal arch	Previous treatment of the target lesion, obstruction caused by SFA aneurysmal disease or dissections	1:1 DCB + Supera vs Supera	Legflow DCB	12	1 / 2
Gouëffic²³ 2019	MC, non-blinded	181	18	RC 2–5, de novo SFA and/or P1 lesion ≥2 and ≤14 cm long, RVD 4–7 mm	Asymptomatic lesion, restenosis, no atheromatous disease	1:1 DES vs BMS	Zilver PTX	24	0 / 3
Haddad¹² 2017	SC, NS	93	100	RC ≥4, stenosis or occlusion ≥30 mm in at least 1 tibial artery	Allergy to paclitaxel and contraindication to DAPT	1:1 DCB vs BA	Luminor 14	12	10 / 6
Liistro¹⁴ (BTK) 2013	SC, non-blinded	132	100	RC ≥4, stenosis or occlusion ≥40 mm of at least 1 tibial vessel with runoff to the foot	Allergy to paclitaxel, contraindication to DAPT, planned major amputation	1:1 DCB vs BA	IN.PACT Amphirion	12	5 / 3
Liistro¹³ (SFA) 2013	SC, non-blinded	104	75	De novo SFA or P1 stenosis ≥50% and ≥40 mm in length; no disease between the SFA lesion and CFA and between the P1 and TPT; ≥1 patent tibial vessel with distal runoff	Contraindication to DAPT; need for major amputation at enrolment, failure to recanalize intended BTK arteries in CLTI patients at risk of major amputation	1:1 DCB + BMS vs BA + BMS	IN.PACT Admiral	12	2 / 1
Mustapha⁹ 2019	MC, single-blinded	442	91	RC 3–5, ≥70% stenosis of ≤2 native arterial paths below the tibial plateau and above the tibiotalar joint; target vessel diameter 2–4 mm, total treated length <320 mm, target lesion not previously stented and at least 20 mm from any stent, a patent inflow artery from the aorta, target vessel reconstitutes at or above the ankle with inline flow to at least 1 patent inframalleolar vessel	Severe comorbidities, hx of stroke, gangrene beyond the toe, ischemic ulceration >4 cm, neurotropic, pressure or other ulcer potentially involving the calcaneus, OM, systemic infection, planned major amputation, prior major amputation, GFR ≤30, significant inflow disease, acute limb ischemia, ISR, thrombus in the target vessel, intended use of adjunctive primary treatment modalities	2:1 DCB vs BA	Lutonix	6	14 / 6
Scheinert²⁰ 2015	MC, single-blinded	60	17	Single or sequential de novo or restenotic ≥70% stenoses 30–200 mm long in the FP segment, RVD 3–7 mm, unobstructed inflow, RC 2–5, ≥1 patent crural vessel with runoff to the foot	Target vessel thrombosis or prior bypass surgery or stenting, severe calcification, planned amputation of the target limb, ≥50% distal lesions	1:1 DCB vs BA	Passeo-18 Lux DCB	12	0 / 2
Spreen¹⁵ 2017	MC, non-blinded	137	100	RC ≥4, IP lesions with ≥50% stenosis, ≥90 mm long, RVD 2–6 mm, unobstructed inflow, outflow ≥1 crural vessel with unobstructed runoff to the ankle joint	Ipsilateral lower limb aneurysm, existing stent, planned/previous surgical revascularization of the limb within 30 days, allergy to contrast or DAPT, hx of HIT, severely calcified lesions, untreatable inflow disease, no distal runoff to the index site	1:1 DES vs BA ± BMS	TAXUS Liberte	60	46 / 40
Zeller¹⁶ 2014	MC, single-blinded	358	100	RC ≥4, RVD 2–4 mm, single or multiple lesions with ≥70% stenosis in one or more main afferent crural vessels including TPT with a cumulative length ≤100 mm, at least 1 patent crural vessel with runoff to the foot, only 1 target lesion per patient	Planned major index limb amputation, untreated inflow lesions (DES and/or DEB not allowed for the treatment of inflow lesions)	2:1 DCB vs BA	IN.PACT Amphirion	60	74 / 45
Zeller¹⁷ 2015	MC, single-blinded	72	78	Single or sequential de novo or restenotic lesions (≥70% stenosis) excluding ISR in up to 2 IP arteries ≥30 mm long, RVD 2–4 mm, unobstructed inflow, ≥1 patent crural vessel with runoff to the foot	Lesions extending beyond the ankle joint, acute target vessel thrombosis, planned major amputation of the target limb, previous bypass surgery or stenting of the target vessel/lesion	1:1 DCB vs BA	Passeo-18 Lux DCB	12	2 / 2

Abbreviations: BA, balloon angioplasty; BMS, bare metal stent; BTK, below the knee; CFA, common femoral artery; CLTI, chronic limb-threatening ischemia; CTA, computed tomography angiography; DAPT, dual antiplatelet therapy; DCB, drug-coated balloon; DES, drug-eluting stent; GFR, glomerular filtration rate; HIT, heparin-induced thrombocytopenia; IP, infrapopliteal; ISR, in-stent restenosis; MC, multicenter; MRA, magnetic resonance angiography; NS, not specified; OM, osteomyelitis; P1, proximal popliteal artery; PSVR, peak systolic velocity ratio; PTX, paclitaxel; RC, Rutherford category; RVD, reference vessel diameter; SC, single-center; SFA, superficial femoral artery; TPT, tibioperoneal trunk.

The total number of patients in the studies (2213) differs from the 1450 patients in the meta-analysis owing to the studies that had <100% CLTI patients in their population. For the studies with <50% CLTI patients, just the number of patients with CLTI was used; however, for those studies with >50% CLI patients, the total study population was used in the analysis.

Patients and Lesions

The studies included 2213 patients. After removing patients without CLTI from studies that contained <50% in their samples, 1450 patients were randomized to a paclitaxel-coated device (n=866) or an uncoated control (n=584), of which 1367 (94.3%) patients had CLTI (range 10–429). Patient baseline characteristics are detailed in Table 2. Mean age ranged from 67 to 76 years and over two-thirds of the patients were men (69%). Diabetes, hypertension, and dyslipidemia were prevalent in most trials. Six studies^9,12,14-17 included below-the-knee lesions, and 5 studies^13,18-20,23 included lesions above-the-knee. There was no association between lesion location and mortality risk.

Table 2.

Baseline Patient Characteristics.

Author	N	Men	SMK	DM	HTN	CD	DLP	Stroke	RC		Lesion				Antiplatelet Therapy
Author	N	Men	SMK	DM	HTN	CD	DLP	Stroke	RC		Type	Location	Length, mmPTX / control		Antiplatelet Therapy
Dake¹⁸ 2016	474	307	404	217	404	41	346	50	2–6	9	DN, RS	FP	63.2±40.5	66.4±38.9	Clopidogrel (75 mg/d) or ticlopidine therapy ≥60 days, lifelong ASA
DeBoer¹⁹ 2017	160	102	79	47	—	—	—	—	2–6	16	DN	SFA	158.0±74.0	158.0±76.0	Clopidogrel 75 mg/d for 3 mo and ASA 100 mg/d indefinitely
Gouëffic²³ 2019	181	122	48	62	109	61	114	—	2–5	18	DN	FP	70.0±40.0	65.0±35.0	—
Haddad¹² 2017	93	-	68	89	79	—	85	—	4–6	100	—	BTK	—	—	DAPT for 12 weeks and ASA 100 mg/d thereafter
Liistro (BTK)¹⁴ 2013	132	106	20	132	98	22	39	12	4–6	100	DN	BTK	129.0±83.0	131.0±79.0	DAPT for at least 4 weeks, then ASA 100 mg/d
Liistro (SFA)¹³ 2013	104	73	53	77	92	39	60	20	3–6	75	DN	SFA	94.0±60.0	96.0±69.0	DAPT with a 300-mg loading dose of each drug at least 12 hours before the procedure
Mustapha⁹ 2019	429	304	259	310	412	220	341	86	3–5	90	DN	BTK	111.8±92.6	94.7±85.4	DAPT, ticagrelor (180 mg/d), or prasugrel (5–10 mg/d) for ≥1 mo and 75–100 mg/d ASA thereafter
Scheinert²⁰ 2015	60	34	41	20	44	19	37	12	2–5	17	DN, RS	FP	51.4±47.2	68.5±57.0	DAPT for 1 mo or 3 mo if bailout stenting performed
Spreen¹⁵ 2017	137	96	63	77	—	52	—	25	4–6	100	DN, RS	BTK	23.1±21.8	21.2±19.3	Carbasalate calcium 100 mg/d indefinitely and 75 mg/d clopidogrel for ≥6 mo
Zeller¹⁶ 2014	358	266	52	263	320	101	255	—	3–6	100	DN, RS	BTK	100.2±90.1	120.9±90.5	—
Zeller¹⁷ 2015	72	57	40	48	62	30	49	—	2–5	78	DN, RS	BTK	113.1±88.1	115.0±86.9	DAPT for 1 mo (ASA 100–325 mg/d and clopidogrel 75 mg/d) or for 3 mo if bailout stenting performed

Abbreviations: ASA, aspirin or acetylsalicylic acid; BTK, below the knee; CD, cardiac disease; CLTI, chronic limb-threatening ischemia; DAPT, dual antiplatelet therapy; DLP, dyslipidemia; DM, diabetes mellitus; DN, de novo; FP, femoropopliteal; HTN, hypertension; PTX, paclitaxel; RC, Rutherford category; RS, restenotic; SFA, superficial femoral artery; SMK, smoking.

Five studies^{9,13,14,19,23} enrolled de novo lesions only; 1 study¹² did not indicate the nature of the lesions, and the remaining^15-18,20 studies included participants with both de novo and restenotic lesions. No details regarding previous interventions or exposure to paclitaxel were given by any study.

Device and Intervention

Eight studies^{9,12-14,16,17,19,20} used paclitaxel-coated balloons; the remainder used paclitaxel-eluting stents^15,18,23 (Table 1). The paclitaxel-coated balloons included Passeo-18 Lux DCB (BIOTRONIK, Berlin, Germany),^17,20 Luminor 14 (iVascular, Barcelona, Spain),¹² Legflow (Cardionovum, Bonn, Germany),¹⁹ Lutonix (BD Bard, Covington, GA, USA),⁹ IN.PACT Amphirion (Medtronic, Minneapolis, MN, USA),^14,16 and IN.PACT Admiral (Medtronic).¹³ The paclitaxel-coated stents included the TAXUS Liberte (Boston Scientific Corporation, Marlborough, MA, USA)¹⁵ and the Zilver PTX (Cook Medical, Indianapolis, IN, USA).^18,23 The IN.PACT Amphirion and Admiral balloons^14,16 delivered the highest dose of paclitaxel at 3.5 μg/mm² and the TAXUS Liberte and Zilver PTX stents^15,23 delivered the lowest doses of paclitaxel at 1.0 and 0.37 μg/mm², respectively. In 1 study,⁹ the devices contained a paclitaxel dose of 2 μg/mm². All other devices delivered a paclitaxel dose of 3 or 3.5 μg/mm².^12-14,16-19 The device diameters used ranged from 2.25 to 10 mm and the lengths between 8 to 220 mm. For most paclitaxel-coated balloon studies, the devices were inflated for at least 120 seconds; however, a minimum time of 30 seconds was considered acceptable in 1 study.20

In 5 studies^{9,13,14,19,23} the paclitaxel-coated group underwent predilation prior to the deployment of the drug-coated device. In the remainder^{12,15-17,19,20} the decision to predilate was left to the physician’s discretion; however, the majority of patients in the paclitaxel group received predilation. Nine studies^{9,12,13,15-19,23} included a predefined antiplatelet regimen as part of their protocol, with most including both clopidogrel and aspirin postoperatively for a determined time period.^{12-14,17,19,20} Two studies^16,23 failed to specify a postoperative antiplatelet regimen. One study used carbasalate calcium combined with clopidogrel as their antiplatelet regimen.¹⁵

Follow-up and Clinical Outcome

The duration of follow-up ranged from 6 to 60 months (mean 25.6), and a follow-up period of at least 12 months was reported in 10^12-20,23 of the 11 studies (Table 1). The authors of 2 studies provided long-term follow-up data that was accepted for publication but in press at the time of the review.^16,23 Overall, mortality rates in this CLTI population were 6.5%, 10.8%, 27.8%, 39.7%, and 40.6% at 6, 12, 24, 36, and 60 months, respectively. There were 161 (18.6%) deaths among 866 subjects in the paclitaxel device group and 116 deaths among 584 (19.9%) subjects in the non-coated control group (RR 0.93, 95% CI 0.78 to 1.12, p=0.45). There was no significant heterogeneity between studies (I²=0%, p=0.58; Figure 2A) nor was there visual asymmetry of the funnel plot to suggest publication bias (Figure 2B). Of the 4 studies^12-14,23 that presented the cause of death, the leading cause was cardiac. The other reported causes of death included stroke, sepsis, fall, respiratory failure, and stroke.^12-14,23 It was not possible to attain further details regarding the specific intervention assigned to each of the patients who died.

Incremental Time Point Analysis

Six studies^{9,15-17,20,23} reported all-cause mortality at 6 months for a total of 1036 subjects. Forty-two (6.5%) of 651 patients in the paclitaxel group died compared to 25 (6.5%) of 385 patients in the control arm (RR 1.06, 95% CI 0.65 to 1.72, p=0.82; Figure 3A). There was no statistical evidence of heterogeneity among studies (I²=0%, p=0.77).

Ten studies^12-20,23 reported all-cause mortality at 12 months for 992 subjects. Of 571 patients in the paclitaxel group, 62 (10.9%) died compared to 45 (10.7%) of 421 patients in the control arm (RR 1.07, 95% CI 0.74 to 1.52, p=0.73: Figure 3B). There was no statistical evidence of heterogeneity among studies (I²=0%, p=0.65).

For the 24-month time point, 3 studies^15,18,23 reported all-cause mortality for 212 subjects. Twenty-seven (22.9%) of 118 patients in the paclitaxel group died compared to 32 (34.0%) of 94 patients in the control arm (RR 0.64, 95% CI 0.31 to 1.32, p=0.23; Figure 3C). There was statistical evidence of low heterogeneity among studies (I²=24%, p=0.27).

Two studies^15,18 reported all-cause mortality at 36 months for 179 subjects. Of 100 patients in the paclitaxel group, 35 (35.0%) died compared to 36 (45.6%) of 79 patients in the control arm (RR 0.76, 95% CI 0.45 to 1.28, p=0.30; Figure 3D). There was moderate statistical evidence of heterogeneity among studies (I²=27%, p=0.24).

At longer follow-up of 60 months, there were 3 studies^15,16,18 and 537 patients for analysis; data from one of the studies was provided by the authors.¹⁶ Of 339 patients in the paclitaxel arm, 127 (37.5%) died vs 91 (46.0%) from 198 in the control arm (RR 0.91, 95% CI 0.75 to 1.10, p=0.32; Figure 3E). Again there was no evidence of statistical heterogeneity (I²=0%, p=0.42).

Methodological Quality and Risk of Bias

The risk of bias assessment for each of the included studies is detailed in Figure 4. Most of the studies had a high risk of bias in domain 2 due to lack of blinding of participants and investigators. The risk of bias arising from the randomization process and from measurement of the outcome was low for each of the studies. There was a degree of missing data for several studies; however, the missing data was minimal and unlikely to affect the results (rated “unknown”). The random sequence generation and selective reporting bias were both considered low risk for most of the studies.

Sensitivity and Subgroup Analyses

Results from the sensitivity and subgroup analyses are given in Table 3. Trials were separated and analyzed according to the dose of paclitaxel coating. The lowest paclitaxel dose was 0.37 μg/mm²,¹⁵ however; this did not correlate with the study with the lowest number of deaths. Similarly, the 3 studies^13,14,16 that administered the highest paclitaxel dose (3.5 μg/mm²) did not correlate with the highest overall mortality risk. Other studies used a paclitaxel dose of 2-3 μg/mm²,^9,12,17-20 which was not associated with either an increased or decreased risk of mortality. The study¹⁹ with the longest lesion length for both interventional groups (158±74.0 mm paclitaxel, 158.0±76.0 mm control) did not correlate with the highest mortality risk, and the study¹⁵ with the shortest lesion length (23.1±21.8 mm paclitaxel, 21.2±19.3 mm control) did not correlate with the lowest mortality risk. One study did not disclose lesion length.¹²

Table 3.

Subgroup and Sensitivity Analyses for All-Cause Mortality.^a

Paclitaxel dose, μg/mm²
<2.0	1.01 (0.78 to 1.31)
2.0–3.0	0.90 (0.54 to 1.49)
3.5	0.86 (0.64 to 1.14)
Device type
Drug-coated balloon	0.91 (0.70 to 1.17)
Drug-coated stent	0.82 (0.45 to 1.48)
Lower limb
Above the knee	0.61 (0.31 to 1.20)
Below the knee	0.96 (0.80 to 1.16)
Lesion length, cm
<10	0.80 (0.45 to 1.41)
≥10	0.87 (0.66 to 1.14)

Data are given as the risk ratio (95% confidence interval).

Seven studies^9,16-20,23 randomized patients, while 4 studies^12-15 randomized lesions. In 3 studies,^12-14 patients found to have >1 lesion were treated with the same intervention as their originally assigned treatment group. The remaining study¹⁵ that randomized lesions did not disclose if patients with >1 lesion were treated with the same intervention as their originally assigned treatment arm; thus crossover into the alternative intervention group may have occurred.

Discussion

This study was performed to assess that single endpoint, mortality, in a patient population with CLTI treated with paclitaxel-coated devices. This effort was in direct response to a meta-analysis that identified higher mortality rates in PAD patients with claudication who underwent endovascular treatment of the lower extremity with paclitaxel-coated balloons and stents.¹⁰ In contrast to those findings, this study found no such association for patients undergoing treatment for CLTI, with similar overall death rates of 18.6% and 19.9% in the drug-coated and uncoated device control groups, respectively. We also observed no effect of paclitaxel dose, target vessel location, or length of the lesion on mortality rates.

Compared to claudicants, patients with CLTI have more advanced multisystem disease, a higher risk of limb loss, and a shorter life expectancy,² illustrated by our finding of a 5-year mortality of 40.6%. Repeat interventions are more frequently required, and each carries a greater burden of risk to both life and limb. An endovascular approach to revascularization has several advantages related to its minimally invasive nature and ease of repeatability that make it an attractive option in this high-risk population where each repeat percutaneous intervention carries significant risk due to concomitant cardiovascular, renal, and respiratory disease.^4-6 It is common to use an endovascular-first approach, reserving open surgery for repeat endovascular failure. However, restenosis and the need for repeat interventions remains an issue with conventional angioplasty balloons and bare metal stents. It is in that context that paclitaxel-coated devices have become an increasingly favored therapy, as their antiproliferative properties have been shown to reduce restenosis,^7,8,24 rates of reintervention,^7,8,24 and healthcare costs.^4,5 These benefits have the potential to translate to improved patient outcomes and systematic efficiencies.^4,25

Paclitaxel has been in continuous use for oncological indications since first registered with the US Federal Drug Agency in 1992.²⁶ For those purposes it is administered as an intravenous infusion at doses several hundred-fold greater than on a drug-coated balloon or stent.^27,28 The drug has been studied extensively in disease states such as curative breast cancer,^28-30 where it has been found to reduce mortality, casting doubt as to the safety concerns raised by the meta-analysis published in the Journal of the American Heart Association.¹⁰ Further doubts have been raised as detailed analysis has failed to identify a predominant cause of death, mechanism of action, or association between dose and mortality³¹ as one would expect from a toxicity phenomenon. Furthermore, several large real-world studies have been unable to replicate the same association between paclitaxel and mortality in PAD or the treatment of the dysfunctional arteriovenous fistula.^32-39

There is a great deal of ongoing work to determine whether the safety signal observed in patients with claudication represents a drug toxicity effect or a statistical association. However, the benefits of paclitaxel-coated devices are well established. Withholding their use in CLTI patients is likely to detrimentally affect patient outcomes as patients are subjected to increased numbers of interventions and high-risk open surgical procedures, likely leading to increasing rates of amputation and mortality. This subset of PAD patients has the most to lose if paclitaxel-coated devices were to become unavailable, and they warrant individual consideration in light of these findings, which suggest there is no increase in all-cause mortality.

Limitations

This present analysis has several limitations that deserve to be acknowledged and discussed. First, the number of studies and participants in this review was smaller than that of the meta-analysis published by Katsanos et al.¹⁰ Therefore, this analysis may be underpowered and susceptible to type II error, particularly at the time points beyond 1 year. Second, several studies that met the inclusion criteria were excluded as the raw data for the CLTI subjects was not provided by the authors, after multiple requests. Third, like the meta-analysis that preceded it,¹⁰ this investigation was performed at a summary level that was unable to collect patient level data and specific causes of death. Further insights are likely to come from a more detailed meta-analysis performed at an individual patient level. Finally, atherosclerosis is a systemic disease, and it is common that patients with CLTI will have paclitaxel-coated devices used in other vascular territories, such as the other leg or the coronary arteries. This is true of both meta-analyses and is likely to have introduced bias to the results.

Conclusion

This meta-analysis has demonstrated that there is no increased risk of all-cause mortality in a predominately CLTI patient population treated with paclitaxel-coated vs uncoated devices. With clear benefit and no suggestion of a link between the use of paclitaxel-coated devices and mortality, we recommend their continued use in this high-risk patient population.

Footnotes

Appendix

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Shannon D. Thomas is a consultant for Abbott Vascular, Medtronic, and Philips Healthcare. Andrew Holden is a medical advisory board member for Gore, Boston Scientific, and Medtronic and a clinical investigator for Cook, BD Bard, Shockwave, Intact Medical, Surmodics, Cagent, and Trireme Medical. Peter A. Schneider is a medical advisory board member for Medtronic, Boston Scientific, and Philips and a consultant for Surmodics, Silk Road Medical, Medtronic, CSI, Profusa, Intact Vascular, Cagent, CTI, Illuminate, Intervene, Limflow, and Devoro; Ramon L. Varcoe is a consultant for Abbott Vascular, Medtronic, Boston Scientific, Intact Medical, Shockwave, Clinlogix, Surmodics, and Intervene.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Shannon D. Thomas

Andrew Holden

Ramon L. Varcoe

References

Song

Rudan

Zhu

, et al. Global, regional, and national prevalence and risk factors for peripheral artery disease in 2015: an updated systematic review and analysis. Lancet Glob Health. 2019;7:e1020-e1030.

Teraa

Conte

Moll

, et al. Critical limb ischemia: current trends and future directions. J Am Heart Assoc. 2016;5(2). doi: 10.1161/JAHA.115.002938.

TASC Steering Committee, Jaff

White

, et al. An update on methods for revascularization and expansion of the TASC lesion classification to include below-the-knee arteries: a supplement to the Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II). J Endovasc Ther. 2015;22:663-677.

Katsanos

Geisler

Garner

, et al. Economic analysis of endovascular drug-eluting treatments for femoropopliteal artery disease in the UK. BMJ Open. 2016;6:e011245.

Salisbury

Vilain

, et al. Cost-effectiveness of endovascular femoropopliteal intervention using drug-coated balloons versus standard percutaneous transluminal angioplasty: results from the IN.PACT SFA II trial. JACC Cardiovasc Interv. 2016;9:2343-2352.

Tang

Paravastu

Thomas

, et al. Cost analysis of initial treatment with endovascular revascularization, open surgery, or primary major amputation in patients with peripheral artery disease. J Endovasc Ther. 2018;25:504-511.

Katsanos

Kitrou

Spiliopoulos

, et al. Comparative effectiveness of plain balloon angioplasty, bare metal stents, drug-coated balloons, and drug-eluting stents for the treatment of infrapopliteal artery disease: systematic review and Bayesian network meta-analysis of randomized controlled trials. J Endovasc Ther. 2016;23:851-863.

Katsanos

Spiliopoulos

Paraskevopoulos

, et al. Systematic review and meta-analysis of randomized controlled trials of paclitaxel-coated balloon angioplasty in the femoropopliteal arteries: role of paclitaxel dose and bioavailability. J Endovasc Ther. 2016;23:356-370.

Mustapha

Brodmann

Geraghty

, et al. Drug-coated vs uncoated percutaneous transluminal angioplasty in infrapopliteal arteries: six-month results of the Lutonix BTK trial. J Invasive Cardiol. 2019;31:205-211.

10.

Katsanos

Spiliopoulos

Kitrou

, et al. Risk of death following application of paclitaxel-coated balloons and stents in the femoropopliteal artery of the leg: a systematic review and meta-analysis of randomized controlled trials. J Am Heart Assoc. 2018;7:e011245.

11.

Moher

Shamseer

Clarke

, et al. Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev. 2015;4:1.

12.

Haddad

Shishani

Qtaish

, et al. One year primary patency of infrapopliteal angioplasty using drug-eluting balloons: single center experience at King Hussein Medical Center. J Clin Imaging Sci. 2017;7:31.

13.

Liistro

Grotti

Porto

, et al. Drug-eluting balloon in peripheral intervention for the superficial femoral artery: the DEBATE-SFA randomized trial (drug eluting balloon in peripheral intervention for the superficial femoral artery). JACC Cardiovasc Interv. 2013;6:1295-1302.

14.

Liistro

Porto

Angioli

, et al. Drug-eluting balloon in peripheral intervention for below the knee angioplasty evaluation (DEBATE-BTK): a randomized trial in diabetic patients with critical limb ischemia. Circulation. 2013;128:615-621.

15.

Spreen

Martens

Knippenberg

, et al. Long-term follow-up of the PADI trial: percutaneous transluminal angioplasty versus drug-eluting stents for infrapopliteal lesions in critical limb ischemia. J Am Heart Assoc. 2017;6. doi: 10.1161/JAHA.116.004877.

16.

Zeller

Baumgartner

Scheinert

, et al. Drug-eluting balloon versus standard balloon angioplasty for infrapopliteal arterial revascularization in critical limb ischemia: 12-month results from the IN.PACT DEEP randomized trial. J Am Coll Cardiol. 2014;64:1568-1576.

17.

Zeller

Beschorner

Pilger

, et al. Paclitaxel-coated balloon in infrapopliteal arteries: 12-month results from the BIOLUX P-II randomized trial (BIOTRONIK’S-First in Man study of the Passeo-18 LUX drug releasing PTA Balloon Catheter vs. the uncoated Passeo-18 PTA balloon catheter in subjects requiring revascularization of infrapopliteal arteries). JACC Cardiovasc Interv. 2015;8:1614-1622.

18.

Dake

Ansel

Jaff

, et al. Durable clinical effectiveness with paclitaxel-eluting stents in the femoropopliteal artery: 5-year results of the Zilver PTX randomized trial. Circulation. 2016;133:1472-1483.

19.

de Boer

van den Heuvel

DAF

de Vries-Werson

DAB

, et al. Short-term results of the RAPID randomized trial of the LegFlow paclitaxel-eluting balloon with Supera stenting vs Supera stenting alone for the treatment of intermediate and long superficial femoral artery lesions. J Endovasc Ther. 2017;24:783-792.

20.

Scheinert

Schulte

Zeller

, et al. Paclitaxel-releasing balloon in femoropopliteal lesions using a BTHC excipient: twelve-month results from the BIOLUX P-I randomized trial. J Endovasc Ther. 2015;22:14-21.

21.

Higgins

Altman

Gotzsche

, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. Br Med J. 2011;343:d5928.

22.

DerSimonian

Laird

. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177-188.

23.

Goueffic

Sauguet

Desgranges

, et al. A polymer-free paclitaxel-eluting stent (Zilver PTX) versus a bare-metal stent (Misago) for de novo femoropopliteal lesions (BATTLE): a randomized trial [abstract]. J Am Coll Cardiol. 2019;74(13 Suppl):B62. DOI: 10.1016/j.jacc.2019.08.097

24.

Katsanos

. Drug-eluting stents will show better 5-year patency rates than PTA: pro. Cardiovasc Intervent Radiol. 2014;37:S103.

25.

Albrecht

Waliszewski

Roca

, et al. Two-year clinical outcomes of the CONSEQUENT trial: can femoropopliteal lesions be treated with sustainable clinical results that are economically sound? Cardiovasc Intervent Radiol. 2018;41:1008-1014.

26.

Song

Dumais

. From yew to us: the curious development of taxol. J Am Med Assoc. 1991;266:1281.

27.

Mayer

. Use of paclitaxel as a primary therapy in oncology: outcomes, benefits, adverse events and doses compared to devices. Abstract presented at: VIVA Vascular Leaders Forum (VLF); March 1–2, 2019; Washington, DC, USA.

28.

Shulman

Berry

Cirrincione

, et al. Comparison of doxorubicin and cyclophosphamide versus single-agent paclitaxel as adjuvant therapy for breast cancer in women with 0 to 3 positive axillary nodes: CALGB 40101 (Alliance). J Clin Oncol. 2014;32:2311-2317.

29.

Mamounas

Bryant

Lembersky

, et al. Paclitaxel after doxorubicin plus cyclophosphamide as adjuvant chemotherapy for node-positive breast cancer: results from NSABP B-28. J Clin Oncol. 2005;23:3686-3696.

30.

Henderson

Berry

Demetri

, et al. Improved outcomes from adding sequential paclitaxel but not from escalating doxorubicin dose in an adjuvant chemotherapy regimen for patients with node-positive primary breast cancer. J Clin Oncol. 2003;21:976-983.

31.

Holden

Varcoe

Jaff

, et al. Paclitaxel and mortality: the dose argument is critical. J Endovasc Ther. 2019;26:467-470.

32.

Shishehbor

Secemsky

Varcoe

. Is there a real association between paclitaxel devices and mortality? time to pause and re-evaluate what we know about this statistical finding. J Am Heart Assoc. 2019;8:e012524.

33.

Freisinger

Koeppe

Gerss

, et al. Mortality after use of paclitaxel-based devices in peripheral arteries: a real-world safety analysis [published online ahead of print October 8, 2019]. Eur Heart J. doi: 10.1093/eurheartj/ehz698.

34.

Secemsky

Kundi

Weinberg

, et al. Drug-eluting stent implantation and long-term survival following peripheral artery revascularization. J Am Coll Cardiol. 2019;73:2636-2638.

35.

Donas

Sohr

Pitoulias

, et al. Long-term mortality of matched patients with intermittent claudication treated by high-dose paclitaxel-coated balloon versus plain balloon angioplasty: a real-world study. Cardiovasc Intervent Radiol. 2020;43(1):2-7.

36.

Katsuki

Takahara

Soga

, et al. Mortality risk following application of a paclitaxel-coated stent in femoropopliteal lesions. J Endovasc Ther. 2019;26:593-599.

37.

Secemsky

. Long-term safety of drug-coated devices for peripheral artery revascularization: an updated analysis of Medicare beneficiary data. Abstract presented at: TCT2019; September 25-29, 2019; San Francisco, CA, USA.

38.

Cortese

Alfonso

Pellegrini

, et al. The hypothesis of an increased mortality following paclitaxel coated device use in peripheral vascular interventions (and the emerging era of meta-analysis based evidence) [published online ahead of print November 19, 2019]. Catheter Cardiovasc Interv. doi: 10.1002/ccd.28600.

39.

Dinh

Limmer

Paravastu

, et al. Mortality after paclitaxel-coated device use in dialysis access: a systematic review and meta-analysis. J Endovasc Ther. 2019;26:600-612.