Keywords

Background

Peripheral arterial disease (PAD) affects >200 million people worldwide and contributes to significant lifestyle-limiting claudication and to significant morbidity and mortality [1]. It involves some of the largest conduits in the body and is associated with significant atherosclerotic disease burden. Advances in endovascular therapies offer a minimally invasive approach to revascularization as an alternative to surgery with equal efficacy and lower periprocedural risk and complications [2, 3]. However, these therapies have limited long-term durability and have a high risk of restenosis.

The discovery of balloon angioplasty in 1977 by Andreas Gruentzig revolutionized the field of interventional cardiology as it offered a minimally invasive therapy for the treatment of coronary artery disease [4]. This was followed by the use of bare metal stents (BMS), which were initially developed to treat complications such as dissections and acute vessel closure; however, these were also limited by high rates of in-stent restenosis that affected nearly 30–40% of percutaneous coronary interventions (PCIs) [5]. This led to the discovery of drugs that inhibit smooth muscle proliferation and neointimal hyperplasia, processes which underlie the pathophysiology of restenosis.

Drug-coated balloons (DCBs) were developed as a therapy to address restenosis by the delivery of antiproliferative drugs to the arterial wall upon balloon inflation. DCBs were first described in human use in 1995 by Camenzind et al. with the use of a coil-based dispatch balloon in 22 patients in the coronary arteries [6]. The porous balloon was designed to elute intracoronary heparin (known inhibitor of smooth muscle cells) under low-pressure inflation while maintaining distal perfusion for targeted therapy in an area of restenosis. This was fallen out of favor by the lack of significant inhibition of neointimal proliferation by the low molecular weight heparin, reviparin [7]. Other antiproliferative drugs such as colchicine and methotrexate were tested, however unsuccessful due to high washout rates and inadequate retention [8, 9]. Scheller et al. first described paclitaxel use in coronary arteries as a treatment for in-stent restenosis [10]. In a small study of 52 patients, 1 out of 22 (5%) patients treated with DCB had restenosis as compared to 10/23 (43%) in the uncoated balloon angioplasty group [10]. Subsequent studies also demonstrated efficacy of paclitaxel in the inhibition of coronary artery restenosis [11, 12]. The evolution and subsequent success of drug-coated therapies in the coronary arteries led to their extension into the treatment of peripheral artery disease.

Patency rates in femoropopliteal disease are limited by long lesions (which can sometimes measure between 200 and 300 mm) that are often accompanied by severe calcifications and chronic total occlusions (CTOs). Patency rates also vary depending on the therapy used. Percutaneous transluminal angioplasty (PTA) in focal lesions has patency rates of about 87% at 6 months and 78% at 3 years [13]. However, in longer lesions, this can be as low as 20–33% at 1 year [14, 15]. Data for the use of atherectomy in addition to PTA is limited to small single-center or registry data which demonstrate improved luminal gain and less bailout stenting; however, patency rates still remain around 60–80% at 1 year [16,17,18]. Although stents improve initial patency compared with PTA alone [18], the sustained benefit remains suboptimal, in particular for longer SFA lesions, with 1-year patency rates of 63–81% [19, 20]. Patency rates of contemporary drug-eluting stents (DES) remain at 79–86% and 77–83% at 12 and 24 months, respectively [21, 22], and drop to 72.4% at 5 years [23]. Challenges for stenting in the femoropopliteal space might include the length and calcification of arteries that can contribute to stent underexpansion and other factors including external compression and torsional forces from hip flexion/extension which can contribute to stent deformation and fractures and thereby lower patency.

Drug-coated balloons (DCBs) were developed to offer an attractive alternative therapy for femoropopliteal disease. With the delivery of sustained, localized anti-restenotic therapy upon balloon inflation, these devices avoid residual scaffold and polymer effects that might contribute to inflammatory or hypersensitivity reactions and thereby restenosis. DCBs are semi-compliant balloons covered in an antineoplastic drug, designed to locally deliver the drug upon contact with the vessel wall. At the time of writing of this chapter, DCB technology contains three main components which include currently approved cytostatic therapeutic agent, coating, and an excipient designed for efficient drug delivery to the vessel wall.

Technical and Pharmacologic Considerations of DCBs

The three components of the DCB—drug, coating, and excipient—play a crucial role in ensuring adequate drug uptake and retention, efficient transfer of the drug to the vessel wall, and minimal systemic loss and toxicity. Paclitaxel and sirolimus have been tested for use in DCBs. While both demonstrate rapid uptake and binding, paclitaxel achieves greater tissue concentrations, especially in the presence of contrast agents [24]. Paclitaxel’s lipophilicity, which ensures drug uptake and retention by the vessel wall, wide therapeutic window, and lack of systemic toxicity at the doses adequate to inhibit restenosis, make it a drug of choice for DCBs.

Paclitaxel, a taxane, is an antineoplastic agent that induces the polymerization of elastin and tubulin leading to the formation of nonfunctional microtubules, thereby impairing intracellular signaling, protein secretion, and cell migration [25]. The inhibition of smooth muscle cell migration and endothelial cells occurs at much lower doses compared to those needed to achieve its antineoplastic effects. The lower doses seemingly affect the microtubule assembly leading to inhibition of cell structure, secretory processes, and motility; however, higher doses are necessary for cell apoptosis [26]. The local antiproliferative effects on vascular smooth muscle and extracellular matrix and inhibition of neointimal proliferation and hyperplasia make it an attractive agent for the treatment of de novo and restenotic atherosclerotic lesions. Early studies revealed inhibition of smooth muscle proliferation up to 14 days after a single-dose exposure to paclitaxel [26].

Current DCBs contain paclitaxel doses ranging from 1.3 to 3.5 μg/mm2 with doses as low as 1 μg/mm2 showing effective inhibition of neointimal hyperplasia [27]. Porcine studies have demonstrated that about 10% of the drug dosage is lost before it reaches target site and about 80% is released during balloon inflation [28]. However, only 20% is retained at the target site, while the remainder is lost into the distal circulation. While drug concentration may decline for a few days after delivery, up to 80% in 24 h [29], there is persistent drug concentration at the site even 6 months after therapy [30]. Additionally, studies show persistent inhibition of smooth muscle cells and collagen accumulation up to 90 days after DCB therapy [31]. Freyhardt et al. studied the bioavailability of paclitaxel in the plasma among 14 patients with SFA disease who underwent paclitaxel DCB therapy [32]. A maximum paclitaxel plasma concentration of 40.1 ± 76.6 ng/mL was found immediately after intervention, and within 24 h, the paclitaxel plasma level was below detectable levels in all patients [32]. In an FDA analysis, paclitaxel dose during treatment with current drug-coated balloons and stents corresponded to a dose range of 0.167–3.5 μg/mm2 (maximum total drug load of 0.1–17 mg), while cytotoxic effects were noted to occur when drug concentrations exceeded 135–175 mg/m2 over 3 to 24-h period [33].

There are several forms of paclitaxel in current DCB coatings ranging from amorphous, crystalline, hybrid, microcrystalline, and nanoencapsulations. Crystalline and amorphous are the most common forms and affect the amount of drug retention, drug loss, and, thereby, the pharmacokinetic behavior of the coating. While the amorphous paclitaxel provides a more homogenous coating, it results in a shorter duration of drug retention. In contrast, the crystalline form results in a prolonged retention time and, thereby, higher tissue levels and biologic efficacy, however at the expense of more drug loss. The crystalline formulation was found to have a lower washout rate (88.6% vs. 99.9% with amorphous at 7 days) and thereby higher concentration in the vessel wall at 7 and 28 days after treatment [34].

However, lipophilic substances are not necessarily soluble, a property required for quick release from the balloon and efficient delivery of the drug to the arterial wall. This was circumvented by the use of excipients or substances that prevent paclitaxel washout and enhance its transfer to the vessel wall. Paclitaxel in a contrast agent resulted in a higher local tissue concentration indicating the need for additional compounds to ensure drug transfer [28]. Early studies use iopromide, a contrast agent to enhance drug delivery, retention, and bioavailability by increasing the solubility of paclitaxel [28]. Preclinical data by Albrecht et al. investigated inhibition of restenosis in varying paclitaxel doses and coating with or without contrast medium [35]. Compared to a rate of in-stent restenosis close to 38% in the control group of uncoated balloons, there was a reduction of restenosis to about 18% in all subgroups treated with paclitaxel doses of 330 μg, 480 μg, and 6.4 mg dissolved in iopromide [35]. Polymer-based carriers, resorbable polymers, or non-polymeric agents such as urea, fatty acids, and contrast agents are also used to enhance efficient drug delivery.

Several other compounds have been used in various DCB platforms—including urea, shellac, BTHC (n-butyryl tri-n-hexyl citrate), polyethylene glycol (PEG), and ATEC (acetyl triethyl citrate). BTHC, a hydrophobic excipient used in the Passeo-18 Lux DCB (Biotronik), enhances paclitaxel retention in tissue to nearly 28 days after exposure [36]. Studies [37, 38] have suggested that there is heterogeneity in drug uptake with the use of various excipients. In a study of 45 patients [39], the IN.PACT Falcon DCB (urea-based) was associated with lower late lumen loss (LLL) and diameter stenosis at 6 months compared to the shellac-based DIOR DCB (Eurocor). The urea-based DCB was also associated with a higher in-stent fractional flow reserve (FFR) and lower decrease in neointimal volume at follow-up by optical coherence tomography (OCT) suggesting variations in the inhibition of neointimal hyperplasia and restenosis [39]. Interestingly, the Lutonix 035 and Stellarex excipient were noted to bind very strongly to the drug causing a slower dissolution rate and minimal drug release upon inflation [40]. Current DCBs that are approved for use in the United States along with their properties are summarized in Table 12.1.

Table 12.1 Paclitaxel DCBs currently available for treatment of PAD in the United States
Full size table

Other important characteristics that play a role in drug delivery to the vessel wall include properties of the balloon catheter and vessel wall characteristics. Homogenous drug coating over the balloon catheter is essential to ensure uniformity in vessel wall coverage. Drug delivery catheters have gone through a series of evolution from hydrogel coated to double balloons and iontophoretic balloon catheters. Contemporary DCBs use balloons in which the drug is stored within pleats and folds to minimize drug loss during transfer. Vessel wall and plaque characteristics are also critical determinants of drug absorption. Calcium presents a barrier to paclitaxel absorption as shown by Fanelli et al. In lesions with significant calcium burden, the effect of paclitaxel was significantly lower [41]. In lesions treated with DCB, those with greater than 270 degree arc of calcium on CT angiography had lower primary patency, ABIs, and higher LLL, TLR, and major adverse events compared to those with less than 90 degree arc of calcium suggesting that the effect of paclitaxel was diminished by the presence of calcium [41].

Lastly, the mechanical effect of balloon inflation not just allows for arterial wall stretch and plaque reduction to achieve an acute luminal gain but also facilitates drug delivery and retention. Drug delivery depends on the duration of balloon inflation, usually 30–60 s, after which drug concentrations plateau. The duration of balloon inflations was evaluated by Cremers et al. who found that balloon inflation times of 10 s were equally efficacious as longer inflations in the treatment of restenosis [42]. The study suggested that the drug (5 μg/mm2 coated balloons) was rapidly transferred, within seconds after balloon inflation. In addition, they also demonstrated that doses up to 10 μg/mm2 from two consecutive DCB inflations did not necessarily correlate with increased toxicity such as aneurysm formation or thrombosis [42].

Since drugs are mounted on semi-compliant balloons, adequate lesion preparation is also essential prior to DCB therapy. Pre-dilation is not just recommended for its acute mechanical effects; it is also shown to improve drug penetration into the vessel wall. Adequate pre-dilation minimizes drug loss upon balloon delivery and results in a more uniform expansion, hence more homogenous drug delivery to the lesion [43].

Clinical Applications of DCBs in PAD

Drug-Coated Balloons in Native Femoropopliteal Lesions

Current guidelines for the management of femoropopliteal disease after failed guideline-directed medical and exercise therapy support the use of endovascular therapies (Class IIA, LOE A) [44]. Endovascular intervention is recommended for patients with lifestyle-limiting claudication with hemodynamically significant lesions (Class IIA). Most lesions that contribute to intermittent claudication (70%) include femoropopliteal lesions, while the others include aortoiliac disease. Although treatment approaches may vary depending on the location, length, and lesion characteristic and clinical factors, an endovascular-first approach is usually recommended for iliac and femoropopliteal disease [45]. DCB use in native femoropopliteal disease is generally associated with improved patency rates compared to plain balloon angioplasty (PTA) as evaluated in several clinical trials, summarized in Table 12.2.

Table 12.2 DCB RCTs in native femoropopliteal lesions
Full size table

Drug-coated balloons in PAD were first evaluated in the Local Taxan with Short Time Contact for Reduction of Restenosis in Distal Arteries (THUNDER) trial in 2008 which assessed the efficacy of using DCBs in femoropopliteal disease among 154 patients [46]. Sixty six of these patients, followed up to 5 years, demonstrated restenosis rate of 17% (vs. 54% in PTA) and target lesion revascularization (TLR) of 21% in the DCB group (vs. 56% in the uncoated PTA) [46]. There was no evidence of safety concerns in the 5-year data including aneurysm formation and constrictive fibrosis [47]. There was significantly lower late luminal loss (LLL) in the DCB group compared to the balloon angioplasty group. Furthermore, these results were maintained despite the presence of moderate to severe dissections without stent implantation [56].

Similar results were established in the Femoral Paclitaxel Randomized Pilot (Fem-Pac) trial of iopromide-paclitaxel DCB vs balloon angioplasty of 87 patients [48]. Significantly lower LLL (0.5 ± 1.1 vs. 1.0 ± 1.1 mm; p = 0.031) and TLR (6.7% vs. 33%; p = 0.002) were observed in the DCB cohort compared to uncoated balloon angioplasty [48]. TLR results were maintained up to 18 months of follow-up. The findings of these trials were further confirmed in the Paclitaxel-coated Balloons in Femoral Indication to Defeat Restenosis (PACIFIER) trial which assessed the efficacy of IN.PACT Pacific DCBs [49]. In addition to the primary endpoints of LLL and restenosis, there is a significant reduction in TLR and major adverse limb events at 1 year compared to uncoated angioplasty [49].

Lower-dose DCBs were studied in the Lutonix Paclitaxel-Coated Balloon for the Prevention of Femoropopliteal Restenosis (LEVANT I and II) trial using Lutonix 2 μg/mm2 dosing using a sorbitol-based excipient. While the LEVANT I RCT showed promising results of reduced LLL at 6 months, there was no difference in TLR at 24 months [50]. The study was limited by a significant rate of balloon malfunction (due to a manufacturing defect of twisted balloon folds) that resulted in failed deployments [49]. Hence, the LEVANT II trial involving 476 patients with 2:1 randomization to Lutonix DCB vs PTA was conducted with a larger sample size and greater statistical power [51]. While there was a significant improvement in primary patency (defined as a composite freedom from TLR and binary restenosis) at 12 months, DCB was shown to be non-inferior to uncoated balloon angioplasty with respect to safety endpoints such as freedom from limb-related events and perioperative death (83% with DCB vs 79% in control, p = 0.005 for non-inferiority) [51]. Notably, there was no significant difference in functional outcomes (including change in Rutherford class, walking impairment score, and quality-of-life measures) and freedom from TLR (87.7% vs 83.2%, p = 0.21) which was concerning [51].

Using clinical endpoints such as clinically driven TLR (either due to symptoms or decrease in ABI), the IN.PACT SFA trial demonstrated that primary patency was still significantly higher in the DCB group up to a follow-up duration of 3 years [52]. In the IN.PACT SFA trial, the 3-year patency rate associated with the use of DCB was 69.5% which was significantly higher than PTA (45%) [52]. There was no significant difference between the two groups with respect to functional improvement. Through 5 years of follow-up, the DCB group demonstrated superiority in freedom from clinically driven TLR compared to the PTA group (74.5% DCB vs. 65.3% PTA, p = 0.02) [53]. Although there was no difference in the primary safety composite endpoint between the two groups, there was a statistically nonsignificant trend toward higher all-cause death in the DCB arm (15.8% DCB vs 9.6% PTA, p = 0.156).

Similar paclitaxel dosing with a polyethylene glycol coating was assessed in the Randomized Trial of a Novel Paclitaxel-Coated Percutaneous Angioplasty Balloon (ILLUMENATE) study using the Stellarex DCB (Phillips, Spectranetics). This study met its superiority endpoint over balloon angioplasty with respect to primary patency at 12 months (83.9% DCB vs. 60.6% PTA, p < 0.001) [54]. The DCB was found to be non-inferior compared to PTA (94% DCB vs. 83% PTA) for the safety endpoint (defined as freedom from TLR and major amputations) [54].

Recently, the Ranger (Boston Scientific) low-dose paclitaxel DCB with a novel TransPax coating was approved for use. In the pilot RCT of the Ranger DCB which included 105 patients, there was a significant improvement in primary patency (86.4% vs. 56.4% with PTA) and freedom from TLR at 12 months [55]. Subsequently, the Ranger II SFA (presented at Vascular InterVentional Advances, VIVA 2019) of 376 patients aim to randomize 3:1 treatment with the Ranger DCB vs. PTA. The primary patency (defined by ultrasound in the absence of CD-TLR or bypass of the target lesion) was 94.1% in the DCB arm compared to 83.5% wit PTA (p < 0.01 for non-inferiority). Significant reduction in CD-TLR (16.5% vs. 5.5%) was seen in the DCB group, and importantly, there was no significant difference in mortality between the two groups.

These proof-of-concept RCTs were supplemented by data from large, multicenter registries. The LEVANT SFA Global was a multicenter, prospective registry that enrolled 691 patients in 38 centers from 10 countries treated with Lutonix DCB for a follow-up period of 24 months [57]. Freedom from TLR, amputation and death was 89.3% overall, 88.2% for long lesions, and 84.6% for in-stent restenosis at 24 months. Site-reported primary patency (defined as the onset of patency failure) was 85.4% and 75.6% at 12 and 24 months, respectively. Seventy-six percent of patients had an improvement in at least 1 Rutherford class [57].

The larger IN.PACT SFA Global Study of 1535 patients and 1773 lesions from 64 sites worldwide and included complex femoropopliteal arterial disease (long lesions (≥150 mm, mean lesion length 12.09 cm), chronic total occlusions (CTOs) (35.5%), calcified lesions (68.7%), and in-stent restenotic (ISR) (18%) lesions) [58]. Procedural and angiographic data were measured by an independent core lab analysis. There was a 92.6% rate of freedom from TLR at 12 months using the IN.PACT Admiral DCB. At 36 months, this dropped to 76.9% and was significantly lower in patients with chronic limb-threatening ischemia (CLTI) vs. intermittent claudication (IC). The composite safety endpoint was met in 75.6% of the study population [58]. Additionally, a multivariate analysis predicted increased lesion length, ISR, bilateral disease, hyperlipidemia, CLTI, and vessel diameter <4.5 mm as predictors of CD-TLR [58].

In the ILLUMENATE Global prospective, multicenter registry of 371 patients treated with the Stellarex DCB, the patency (evaluated by Doppler ultrasound) was 81.4% and freedom from CD-TLR 95% at 12/24 months [59]. This was accompanied by clinical improvement in Rutherford classification to 90.3% and walking impairment questionnaire by 84% [59]. At 2 years, however, the primary patency was lower, at 72.4% [60].

Drug-Coated Balloons in Below-the-Knee Interventions

Endovascular interventions are an emerging tool for the management of below-the-knee critical limb-threatening ischemia (CLTI). Given the increased risk of amputation and major adverse cardiovascular events, the ACC/AHA 2018 guidelines provide an appropriate use indication for both endovascular and surgical approaches to the treatment of below-the-knee vessels [61]. Endovascular therapies to establish direct in-line blood flow to the foot in patients with nonhealing wounds or gangrene (Class I) is recommended. However, strong evidence for the use of DCB in below-the-knee interventions is still lacking.

In a meta-analysis of 6769 patients treated with infrapopliteal disease, primary patency at 1 year was 63%, 15% underwent repeat revascularization, and about the same also underwent amputation and died as a result of all-cause mortality [62]. Possible challenges with tibial artery interventions include extensive calcifications; long, diffusely disease segments; large number of CTOs; inability to estimate true vessel size; and, therefore, appropriate balloon sizing and possible loss of drug from the DCB due to longer delivery time.

Several studies have explored the safety and efficacy of DCBs in infrapopliteal disease. Table 12.3 summarizes the major RCTs that assessed the use of DCBs in major RCTs. Most of these trials are limited by a small sample size. The Drug-Eluting Balloon Evaluation for Lower Limb Multilevel Treatment (DEBELLUM) trial [63] which included 50 patients had about 25% below-the-knee lesions. At 6- and 12-month follow-up, the DCB arm had significantly lower LLL (0.5 mm vs 1.6 mm), TLR (6.1% vs. 23.6%), and binary restenosis rates (9.1% vs. 28.9%) compared to PTA. In addition, there was a significant improvement in Fontaine class in the DCB group compared to PTA (80% vs. 56%). The Drug-Eluting Balloon in Peripheral Intervention for Below-the-Knee Angioplasty Evaluation (DEBATE-BTK) trial evaluated patients with long lesions below the knee, and a significant proportion of them had occlusions (77–82%). In spite of these high-risk lesions, there was still a superiority benefit with the use of DCBs with respect to TLR and binary restenosis rates at 12 months [64].

Table 12.3 DCB RCTs in infrapopliteal PAD
Full size table

These smaller trials which were followed by data from a large, multicenter randomized trial in below-the-knee critical limb ischemia [65] failed to show any significant benefit with the use of DCBs over PTA in the tibial arteries. In the Randomized IN.PACT Amphirion Drug-Coated Balloon vs. Standard Percutaneous Transluminal Angioplasty for the Treatment of Below-the-Knee Critical Limb Ischemia (IN.PACT DEEP) trial, 358 subjects were randomized 2:1 to DCB angioplasty with the IN.PACT Amphirion balloon or uncoated balloon angioplasty. The study demonstrated no significant benefit of DCB angioplasty over uncoated PTA at 5 years with respect to restenosis and TLR rates. This study also assessed a safety endpoint of all-cause death up to 5 years. While there was no difference in all-cause death, there was a nonsignificant but 2.4-fold increased risk of major amputations compared to the DCB group at 12 months and at 5 years of follow-up [65].

More recently, the IN.PACT BTK study (presented as an LBCT at TCT Connect 2020) evaluated the efficacy of the IN.PACT 0.014 DCB (3.5 μg/mm2) compared to uncovered balloon angioplasty as a control. The study included patients with critical limb ischemia (CLI) and CTOs of the infrapopliteal arteries and lesions that were severely calcified and long (average length 17.6 cm). The results were promising, as the DCB group was associated with a 53% lower LLL compared to the PTA control group, both in the subsegmental (across the entirety of the lesion) and the classic LLL (at the narrowest segment of the artery) groups. In addition, there was no significant difference in the safety endpoints which included device- and procedure-related death, major amputations, and clinically driven TLR at 9 months.

In a network meta-analysis by Katsanos et al., the secondary endpoint of TLR in infrapopliteal arteries demonstrated a significant 40% reduction associated with DCB use. The crude risk of TLR was 11.8% in the DCB group versus 25.6% in the uncoated balloon group. The calculated pooled risk ratio was 0.53 (95% CI 0.35–0.81; p = 0.004), with a corresponding number needed to treat eight patients (95% CI 4–25) [66].

DCB Use in Other Lesion Subsets: In-Stent Restenosis, Long Lesions, and Combination Therapy

In-Stent Restenosis (ISR)

Outcomes for the treatment of ISR depends on the type of lesion—focal lesions, classified as <5 cm; diffuse, generally >5 cm; and total occlusions [67]. In-stent lesions that are total occlusions are usually associated with high recurrent ISR than focal lesions in spite of the use of adjunctive therapies such as atherectomy [68].

Early studies from a single-center, Italian registry of 38 patients suggested successful treatment of SFA ISR with PTA followed by DCB post-dilation. This study showed a 2-year patency rate (defined as proximal velocity ratio of <2.4) of 70.3% in those treated with DCB [69]. This was followed by the Femoropopliteal In-Stent Restenosis Repair: Midterm Outcomes After Paclitaxel-Eluting Balloon Use (PLAISIR) trial, a small sample size prospective cohort study involving 53 patients with femoropopliteal ISR treated with IN.PACT Admiral DCB, which was found to have a 1-year patency rate of 84% and freedom from TLR of 90.2% [70]. Subsequent studies [71,72,73,74] demonstrated the efficacy of DCB in treating ISR lesions; however, these were limited by single-center studies and small study populations. These findings are reviewed in Table 12.4.

Table 12.4 DCB use in in-stent restenotic lesions
Full size table

A meta-analysis by Cassese et al. included four RCTs which evaluated DCB angioplasty for femoropopliteal ISR [75]. A total of 367 patients were followed for a period of 12 months. The study showed that DCB angioplasty resulted in a lower risk of TLR, recurrent ISR, and sustained improvement in Rutherford class compared to plain balloon angioplasty. The DCB group had a lower risk for TLR (odds ratio 0.20, p = 0.002) and recurrent ISR (OR 0.24, p = 0.003) and a sustained RC improvement (OR 2.57, p = 0.002) compared to PTA [75].

This was also confirmed from a subsequent patient-level meta-analysis (2532 patients and 16 RCTs) which showed superiority of drug-coated therapies in ISR lesions [76]. The analysis compared multiple therapies including bare nitinol stents, covered stents, paclitaxel, or sirolimus stents and paclitaxel balloons with plain balloon angioplasty in femoropopliteal ISR. Restenosis and TLR were the lowest in paclitaxel stents and balloons, respectively, supporting the use of drug-coated therapies in in-stent restenotic lesions [76].

Long Lesions

Long femoropopliteal lesions remain a challenge for endovascular intervention with patency rates of 35–50% and often complicated by restenosis that is difficult to treat [77]. The SFA-Long study specifically addressed the role of DCB therapy in long femoropopliteal lesions [78]. The prospective, multicenter, single-arm study included 105 patients who had lesions greater than 15 cm (average lesion length 251 mm) and followed them for a period of 12 months. At follow-up, the patency rate, defined as freedom from combined endpoints of CD-TLR and by duplex ultrasound, was 83.2% [78]. In addition, at 12 months, there was a significant improvement in the quality of life (measured by walking impairment questionnaire) and in the ankle-brachial indices [78]. There was a 7% risk of adverse events which included death from any cause, thrombosis, or nontarget vessel revascularization [78].

In the single-arm sub-study of 131 patients from the IN.PACT Global study [79] with long and complex native ISR lesions (N = 149 lesions, 59% calcified and 34% CTOs), there was an 88% primary patency rate at 12 months treated with DCB. Preliminary data confirmed the safety and efficacy of using DCBs in long and complex ISR lesions in the femoropopliteal arteries [79]. Clinically driven TLR rate was 7.2%, and freedom from device-/procedure-related events was 92.7% at 1 year [79].

Combination Therapies

The increasing prevalence in the use of DCBs was accompanied with the development of several other therapies to address restenosis and improve patency rates in PAD. Most RCTs evaluating the use of DCBs in PAD excluded the use of other adjunctive therapies such as atherectomy, laser atherectomy, scoring balloons, and stents in their trial subjects. Data regarding the use of these adjunctive therapies with drug-coated balloons is limited. Calcium presents a barrier to paclitaxel absorption as shown by Fanelli et al. which showed significantly lower patency rates in calcified lesions treated with DCB [41]. In addition, elastic recoil, incomplete stent expansion, and dissections might affect the patency rates of balloon angioplasty, especially in long and calcified lesions. Atherectomy appears to reduce the risk of dissections and bailout stenting and improve acute procedural results [80, 81].

The Directional Atherectomy (HawkOne, Silver Hawk, TurboHawk Medtronic) Followed by a Paclitaxel-Coated Balloon to Inhibit Restenosis and Maintain Vessel Patency (DEINITIVE-AR) study first evaluated if vessel preparation with atherectomy prior to drug delivery improved outcomes [82]. It was suggested that plaque modification with directional atherectomy might enable a more homogenous drug delivery and increased penetration into the vessel wall [83]. In this pilot, prospective, multicenter randomized study of 102 patients, subjects were randomized to treatment with DAART (directional atherectomy with anti-restenotic therapy) plus DCB vs. DCB alone (Cotavance paclitaxel balloon with Paccocath coating, 3 μg/mm2) alone [80]. The mean lesion length was 106 mm, and about 28% of them were occlusions [80]. One-year primary outcome of angiographic percent diameter stenosis was 33.6% in the atherectomy + DCB arm vs. 36.4% in the DCB arm (p = 0.48). Primary patency and rates of major adverse events were similar between the two groups demonstrating that the use of directional atherectomy prior to DCB therapy was safe and effective [80]. The rate of flow-limiting dissections was lower in the atherectomy plus DCB arm (2% vs. 19.4%, p = 0.01) [82]. In a single-center study of 78 patients, there was a statistically significant difference between primary patencies at 1 year associated with the use of DAART which was 82% compared to 65% with DCB alone [83].

In the recently presented data from the prospective Directional Atherectomy and DCB to Treat Long Calcified Femoropopliteal Lesions (REALITY study presented at Vascular InterVentional Advances (VIVA) 2020, November 6–7, 2020) study of 102 subjects evaluated the use of TurboHawk (Medtronic) atherectomy in conjunction with DCB. Preliminary data demonstrated a 12-month primary patency rate (by duplex ultrasound) of 77% and freedom from CD-TLR of 93% with the use of directional atherectomy prior to DCB (IN.PACT Admiral DCB) for long, calcified femoropopliteal lesions. The average lesion length was 17.9 cm, 39% were CTOs, and 86% had moderate-severe calcification.

Other debulking techniques, especially in the setting of in-stent restenosis, have also been gaining traction. The use of laser atherectomy (Philips) to debulk the ISR lesion, prior to the application of DCB, might help reduce restenosis and improve patency. While laser atherectomy has been shown to improve patency rates compared to PTA alone in the treatment of ISR [84], few studies have investigated its application with DCB. A dual-center observational study of 112 patients with Tosaka II–III lesions [67] underwent laser atherectomy plus DCB treatment and compared with laser atherectomy and plain balloon angioplasty. The use of DCB was associated with a significantly higher freedom from reocclusion (86.7% vs. 57.1%) and TLR (72.5% vs. 50.5%) at 1 year [85]. Further data from the currently enrolling Photo-Pac trial by Zeller et al. might add additional information regarding this therapy.

The use of orbital atherectomy (OA, CSI) in ex vivo peripheral arteries demonstrated a 50% increase in radiolabeled paclitaxel uptake and deeper penetration compared to the untreated segment in calcified plaque [86]. OA use was associated with thinner intima and less plaque calcification [86]. OA with the use of DCB was also investigated in a small retrospective study of 113 patients by Kokkinidis et al. [87]. In patients with heavily calcified femoropopliteal lesions, there were similar outcomes at 2 years with the use of OA in addition to DCB treatment compared to DCB alone (p = NS). The 2-year freedom from TLR was 76.1% in the OA plus DCB group vs. 55.5% in the DCB alone group [87]. There was a significantly lower rate of bailout stenting in the OA group; however, other procedural outcomes and complications were similar between the two groups.

More recently, the use of Jetstream atherectomy (JA) device (Boston Scientific), a rotational cutter with aspiration capacity approved for treatment of calcified femoropopliteal disease, was evaluated in adjunction to DCB in the single-center JET-SCE comparative study of 75 patients [88]. At 16 months, there was a significantly higher freedom from TLR associated with JA and DCB use compared with JA plus PTA (94.4% vs. 54%) [88]. The currently enrolling JET-RANGER randomized study will provide additional information.

The use of drug-eluting stents with DCB remains controversial. While the use of two different antineoplastic agents (paclitaxel-based balloons with limus-based stents) might offer a potential option in complex or recurrent ISR lesions, there remains concern about drug toxicity in overlap areas, incomplete endothelialization of stent struts, and the need for longer duration of dual antiplatelet agents [89].

Future Innovations in DCB Technology

A newer-generation, paclitaxel-coated SurVeil® DCB (Surmodics Inc.) was studied in PREVEIL, a prospective, US, multicenter, feasibility trial at three different clinical sites for the treatment of native femoropopliteal arteries. The lower-dose paclitaxel DCB (2 mcg/mm2 loading dose) uses an excipient that improves efficacy and uniformity of paclitaxel drug transfer and minimizes systemic doses. In total, 13 patients were included with an average lesion length of 56 mm (Vascular InterVentional Advances (VIVA) 2018, Las Vegas, NV). Median paclitaxel plasma concentration peaked immediately post-procedure (Cmax 1.07 ng/mL) and was undetectable at 30 days (Vascular InterVentional Advances (VIVA) 2018, Las Vegas, NV). The plasma concentration achieved was much lower than the currently available DCBs, and there were no adverse events related to the drug or device reported. In addition, there was an improvement in ABI by 0.28, improvement in 6-min walk test by 90.4 m, and improvement in Rutherford class by 69%. Primary patency was achieved in all patients, LLL was 0.27 mm, and there were no reported TLR at 6 months (Vascular InterVentional Advances (VIVA) 2018, Las Vegas, NV). Although not available for clinical use in the United States, further larger trials including TRANSCEND RCT that compares the SurVeil DCB to IN.PACT DCB are still pending.

While sirolimus- and everolimus-coated stents have demonstrated safety and efficacy in the treatment of femoropopliteal disease [90, 91], limus-coated balloons offer a new alternative therapy for de novo and restenotic lesions. Given that limus-based, coronary stents have been shown to have more antiproliferative effect than paclitaxel-based stents [92] and are the cornerstone of interventional therapies for coronary artery disease, limus-based DCBs might offer the same benefit. Sirolimus is a cytostatic molecule that binds to the FKBP-12 molecule blocking cell cycle progression from G1/S phase [93]. However, given short tissue retention time and slow absorption, delivery of sirolimus for several weeks might be necessary for effective inhibition or neointimal proliferation [94]. The physical properties of the drug have made it difficult to be applied in DCB technology. Three sirolimus-coated balloons have been developed for intracoronary use—Magic Touch (Concept Medical) with a phospholipid excipient and carrier, the Virtue balloon (Orchestra BioMed), and SELUTION (MedAlliance). These balloons are also being applied in the peripheral vasculature.

The pilot SELUTION trial was a prospective, multicenter, single-arm trial that assessed 6-month safety and efficacy outcomes of the SELUTION SLR DCB in femoropopliteal arteries [95]. In this study of 50 subjects, outcomes including angiographic late lumen loss (LLL), binary restenosis, improvement in Rutherford class, ankle-brachial index, and patency by duplex ultrasound were all in favor of DCB over plain balloon angioplasty. The patency rate, measured by duplex ultrasound, was 88.4%, and freedom from binary restenosis was 91.2% at 6 months with the use of DCB [95]. Longer-term data pertaining to the use of this DCB remains to be seen. The ongoing XTOSI study by Edward Choke aims to study the application of the Magic Touch, a sirolimus-coated DCB in femoropopliteal arteries in critical limb ischemia.

Paclitaxel scoring balloons also offer an exciting new frontier in the treatment of PAD. With the ability for simultaneous plaque modification and drug delivery, these devices might offer better patency and restenosis rates compared to standard therapy. The recently approved Chocolate Touch DCB balloon was evaluated in the ENDURE trial which included 67 patients, majority with Rutherford class III. The 6-month patency rate was 90% by Duplex ultrasound and LLL of 0.16 mm, which was improved compared to the Lutonix and Stellarex DCBs [90]. Final results remain to be seen.

Mortality and Paclitaxel

In December 2018, a study-level meta-analysis by Katsanos et al. [96], which pooled data from 28 RCTs and included 4432 patients, found an increase in mortality associated with DCB use. While there was no statistically significant difference in 1-year mortality in the 28 RCTs, an increase in 2-year all-cause mortality was reported based on 12 RCTs which persisted at 5 years in 3 studies.

Of the three RCTs (THUNDER, ZILVER-PTX, and IN.PACT SFA) with 5-year follow-up, there was a 14.7% risk of death in the paclitaxel-treated arm vs. 8.1% in the non-paclitaxel arm (6.6% absolute risk) [96]. The authors reported a dose-related increase in mortality postulated to be due to downstream embolization and toxicity. However, the analysis was limited by study-level pooled data and the absence of patient-level data including individual cause of deaths to establish a causal relationship. The study was also limited by several RCTs that were underpowered to detect a mortality difference between the two groups, missing data beyond 2 years, crossover of treatment, and patients lost to follow-up. This led to an FDA guidance to healthcare professionals that cautioned against the use of drug-coated balloons due to an increase in long-term mortality [97].

Following this, a 3-year patient-level meta-analysis of the ILLUMENATE trials [98] was performed and failed to show any association of the Stellarex DCB with mortality for up to 3 years. In the meta-analysis, patient-level data from all patients treated with Stellarex DCB from six studies were included. Of the 1906 patients that were included, all-cause mortality was similar among the groups treated with DCB and PTA [98]. All-cause mortality was 2.1% at 1 year and 7.0% at 3 years of follow-up. Also, an independent patient-level meta-analysis of 1980 patients with 5-year follow-up data of the IN.PACT Admiral DCB showed no association of paclitaxel dose exposure and mortality [99].

Subsequently, an independent FDA analysis included the four RCTs (Zilver PTX, LEVANT 2, IN.PACT SFA I and II, and ILLUMENATE) with available 5-year follow-up data and incorporated trials that were not part of the Katsanos meta-analysis. A patient-level analysis by the FDA also found an increase in mortality associated with paclitaxel use, especially between 2 and 5 years (13.7% vs. 18.3% with DCB); however, this was limited by missing data of 14–38% in some trials [100]. At the same time, the Vascular InterVentional Advances (VIVA) physicians group meta-analysis was performed, which included 2185 subjects from 8 paclitaxel trials with 4-year follow-up of patient-level data and recovered missing data that was not included in the original meta-analysis. This study demonstrated a 4.6% absolute increase in all-cause death in patients treated with paclitaxel devices, which was lower than previously reported [101]. In addition, there was no evidence of drug dose-related exposure and mortality risk over 5 years [101]. However, since freedom from TLR and clinical improvement were still maintained, a full discussion of risks and benefits related to paclitaxel devices was recommended by the FDA.

Following this data, several large observational studies including Vascular Quality Initiative (VQI), OPTUM, and SAFE-PAD attempted to elucidate this mortality signal better. In the propensity-matched, Vascular Surgery VQI analysis of 8376 patients undergoing endovascular treatment of femoropopliteal disease, there was no difference in mortality at 1 year between paclitaxel and non-paclitaxel groups (9.6% vs. 12.6%, respectively) [102]. Mortality was lower in the paclitaxel subgroup of patients with intermittent claudication; however, this was not significant in the CLTI group [102]. A multicenter cohort of 16,560 patients from the Centers for Medicare and Medicaid Services, by contrary, found a lower cumulative incidence of all-cause mortality among those treated with paclitaxel therapies in femoropopliteal arteries (32.5% vs. 34.3%, p = 0.007). After multivariate adjustment, there was no difference in all-cause mortality between the two groups including patients with CLI and among those treated with DCB alone or DES with or without DCB [103]. This was also maintained among patients treated with DES compared to BMS in a Medicare cohort of 51,456 patients (51.7% DES vs. 50.1% BMS at 4.1 years, p = 0.16) [104]. Further data is awaited in the larger yet SAFE-PAD observational retrospective study by Secemsky et al., which includes all Medicare beneficiaries undergoing lower extremity revascularization.

Additional independent analyses that included large registry data in Germany also failed to show any association of mortality related to paclitaxel devices. In a large, propensity-matched cohort of 14,738 patients, there was no evidence of increase in mortality associated with paclitaxel use [105]. In a propensity-matched model, there was a lower all-cause mortality, amputation, and cardiovascular death in the paclitaxel cohort [105]. Freisinger et al. included all patients with BARMER health insurance who underwent any paclitaxel-based therapy for PAD (64,771 patients) [106]. The study also found no association of paclitaxel-coated devices with mortality for up to 11 years posttreatment [106]. Surprisingly, during the first year of follow-up, there was a decrease in mortality with paclitaxel devices compared to uncoated devices (HR 0.92, p < 0.001) which disappeared subsequently [106]. In a retrospective analysis of a similar cohort of 37,914 patients from the BARMER insurance group, a propensity score-matched analysis found an improved overall and amputation-free survival in the paclitaxel-treated CLTI and intermittent claudication (IC) groups [107].

The more recently published results of the SWEDEPAD [108] RCT also support these findings. In this randomized study, 2289 patients were randomized to drug-coated devices vs. uncoated therapies (65% with CLTI) and followed for 2–4 years. The multicenter trial was powered to detect a difference in the primary endpoint of mortality between the two groups. During the follow-up period, there was no difference in all-cause mortality between the two groups. The overall mortality in the DCB group was 10.4%. vs. 9.8% in the uncoated device group [108] which was not significant. The wealth of data since the publication of the meta-analysis by Katsanos et al. continue to support that paclitaxel-coated devices are not associated with an increase in all-cause mortality compared to uncoated devices (Fig. 12.1).

Fig. 12.1
figure 1

Mortality risk associated with paclitaxel use among various contemporary studies. Hazard ratio greater than one corresponds to an increased risk in mortality. Study population size and duration of follow-up are listed along with the referenced study

Full size image

Cost-Effectiveness of Drug-Coated Balloons

In addition to the lower rates of TLR and restenosis associated with DCB therapies, there is also evidence for an economic benefit related to these devices. The economic impact of drug-coated therapies was first examined by Pietzch et al. in a systematic economic analysis that included payers and providers in the United States and Germany [109]. Thirteen studies with 2406 subjects were included following BMS, DCB, DES, and plain balloon angioplasty. The 24-month probability for TLR was 14.3% vs. 19.3% vs. 28.1% vs 40.3% in the DCB, DES, BMS, and PTA groups, respectively. This corresponded to a lower economic budget over 24 months in the drug-coated groups compared to uncoated therapies ($10,214 DCB vs. $12,904 DES vs. $13,114 PTA vs. $13,802 BMS) [109].

Following this data, in a prospective randomized economic study of 181 patients from the IN.PACT SFA II trial, Salisbury et al. evaluated resource utilization over 2 years [110]. While the initial costs were $1129 higher in the DCB cohort, at 24 months of follow-up, limb-related costs were $1212 lower in the DCB group compared to uncovered balloon angioplasty (PTA) [110]. This is attributed to a lower number of repeat revascularization procedures in the DCB group including repeat target limb revascularization, amputation, and recurrent hospitalization, thereby contributing to increased costs in the PTA group. Overall, with the offset of costs between the two groups, there was no significant difference between the two groups at 2 years [110]. This data was similar to that derived from the UK National Health Service [111] which also reflected a lower lifetime costs and greater effectiveness associated with drug-coated therapy.

Conclusion

Since their initial discovery and application, drug-coated technology has made major advances and broadened its application in the treatment of infrainguinal peripheral artery disease. From the treatment of native femoropopliteal arteries to their extension into below-the-knee vessels, in-stent restenosis, and in combination therapies, DCBs play a critical role in the prevention of restenosis. The role of DCB in several of these applications remains to be elucidated as long-term data continues to emerge. While earlier concerns regarding the safety of paclitaxel therapies reflected an increase in mortality, several large, multicenter studies since then have failed to reproduce this finding. Sirolimus-coated balloons also offer an emerging therapeutic alternative for this technology. DCBs continue to remain a critical component for the treatment of PAD with their ability to efficiently deliver therapies to inhibit restenosis and improve patency rates in infrainguinal arterial revascularization.