REVIEWS Drug-coated balloon therapy in coronary and peripheral artery disease Robert A. Byrne, Michael Joner, Fernando Alfonso and Adnan Kastrati Abstract | Nonstent-based local drug delivery during percutaneous intervention offers potential for sustained antirestenotic efficacy without the limitations of permanent vascular implants. Preclinical studies have shown that effective local tissue concentrations of drugs can be achieved using drug-coated balloon (DCB) catheters. Matrix coatings consisting of a mixture of lipophilic paclitaxel and hydrophilic spacer (excipient) are most effective. Clinical applications most suited to DCB therapy are those for which stent implantation is not desirable or less effective, such as in-stent restenosis, bifurcation lesions, or peripheral artery stenoses. Randomized trials have shown superiority of DCBs over plain-balloon angioplasty for both bare-metal and drugeluting coronary in-stent restenosis, and similar efficacy as repeat stenting with a drug-eluting stent (DES). Bycontrast, randomized trials of DCBs in de novo coronary stenosis have, to date, not shown similar efficacy to standard-of-care DES therapy. In peripheral artery disease, DCB therapy has proven superior to plain-balloon angioplasty for treatment of de novo femoropoliteal and below-the-knee disease, and shown promising results for in-stent restenosis. Overall, however, despite many years of clinical experience with DCBs, the number of large, high-quality, randomized clinical trials is low, and further data are urgently needed across the spectrum of clinical indications. Byrne, R. A. et al. Nat. Rev. Cardiol. 11, 13–23 (2014); published online 5 November 2013; doi:10.1038/nrcardio.2013.165
Introduction
Deutsches Herzzentrum, Technische Universität, Lazarettstrasse 36, D‑80636 Munich, Germany (R. A. Byrne, M. Joner, A. Kastrati). Cardiac Department, Hospital Universitario de La Princesa, IIS-IP, Universidad Autónoma de Madrid, c/ Diego de León 62, Madrid 28006, Spain (F. Alfonso). Correspondence to: A. Kastrati [email protected]
The aim of catheter-based percutaneous intervention is the durable relief of arterial obstruction with a mini mal therapeutic footprint. Balloon angioplasty and its forerunner, catheter dilatation (‘dottering’), were pioneering clinical advances. However, the principal limit ations were twofold: modest acute gain and unstable acute results (owing to arterial recoil, plaque prolapse, and vessel dissection), and reduced long-term patency (owing to constrictive remodelling and neointimal hyperplasia). Implantation of metallic or, indeed, polymeric stents results in increased acute gain and stability of acute results, and improved mechanical scaffolding. However, these improvements occur at the expense of increased vessel injury and, consequently, increased levels of neointimal hyperplasia, which is measured in studies as ‘late luminal loss’ at angiographic follow-up.1,2 Drug-eluting stents (DESs) further improve outcomes by combining high mechanical acute gain with local drug delivery to reduce late luminal loss.3,4 The collateral cost here, however, is a systematic delay in healing of the stented vessel, which is associated with a notable increase in late stent thrombosis compared with bare-metal stents, at least with early-generation DES devices.5–7 Competing interests M. Joner declares associations with the following companies: Abbott, Biotronik, Cardionovum, Medtronic, St. Jude Medical, and Zorion. A. Kastrati declares associations with the following companies: Biosensors, Biotronik, and St. Jude Medical. See the article online for full details of the relationships. R. A. Byrne and F. Alfonso declare no competing interests.
Nonstent-based local drug delivery with the use of drug-coated balloons (DCBs) is an alternative to DES technology with broad clinical applicability. The main theoretical advantages over stent-based local drug delivery include a broader area of surface contact and morehomogenous drug–tissue transfer; potential amelioration of delayed arterial healing owing to absence of long-term inflammatory nidus (no requirement for durable polymer surface coatings with DCBs); preservation or early restor ation of normal vessel anatomy and function; and preferential application in scenarios where stent implantation is not desirable (in-stent restenosis, bifurcation carina, or diffuse femoropopliteal artery disease). The limitations of DCB therapy are primarily those shared with balloon angioplasty in the early phase after intervention— namely, lower acute gain and potentially more-unstable acute results. DCB catheters are broadly similar in construction to standard angioplasty catheters (Figure 1). The main distinction is that the balloon is coated with a thin, pharmacologically active surface layer. The primary challenges from a medical engineering standpoint are effective drug transfer within the time frame imposed by a single inflation of the balloon (typically a maximum of 30–60 s in coronary intervention) and the minimization of drug loss during device handling and tracking to the lesion. The surface coating is typically composed of a mixture of fairly high-dose antiproliferative drug (most commonly paclitaxel) and a carrier substance or excipient. Currently available DCB devices and coating constituents
NATURE REVIEWS | CARDIOLOGY
VOLUME 11 | JANUARY 2014 | 13 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Key points ■■ Nonstent-based local drug delivery with drug-coated balloons (DCBs) can achieve rapid transfer of a drug to surrounding tissue, and durable antirestenotic efficacy ■■ Theoretical advantages of DCBs over drug-eluting stents (DESs) include broad surface contact, homogenous drug distribution, absence of stent footprint or polymer residue, and restoration of normal vessel anatomy and function ■■ Preclinical studies have shown that DCB delivery of paclitaxel combined with a hydrophilic spacer (excipient) results in clinically effective local tissue drug concentrations and sustained inhibition of neointimal growth ■■ Among patients with coronary bare-metal or drug-eluting stent restenosis, DCBs show similar results to standard-of-care treatment (repeat stenting with a DES) and obviate the need for further stent implants ■■ Among patients with de novo coronary artery disease, convincing data from randomized trials to support DCB therapy is lacking, and the use of DCBs for this indication requires further investigation ■■ In peripheral artery disease, DCB therapy has proven superior to balloon angioplasty for treatment of de novo femoropopliteal and below-the-knee disease
a
b
c
d
e Balloon surface Matrix coating Endothelium Active drug
Plaque or neointima Tunica media
Exipient
Figure 1 | Drug-coated balloon structure and mechanism of action in in-stent restenosis. a | Neointimal hyperplasia can slowly narrow a stented lesion, causing instent restenosis. b | The drug-eluting balloon is advanced to the lesion site. c | When the balloon is well positioned, the balloon is inflated for at least 30 s, releasing the antiproliferative drug. d | The balloon is withdrawn and the drug penetrates into the artery wall. Paclitaxel will act over several days to minimize cell regrowth. e | In most of these devices, drug release is facilitated by a coating that consists of a mixture of lipophilic drug and a hydrophilic spacer or excipient that facilitates uptake into the vessel wall. The spacer or excipient combines with the active drug to form a matrix coating that remains on the surface of the balloon catheter.
are shown in Table 1. In this Review, we discuss the use of DCBs to treat de novo or in-stent restenosis of coronary or peripheral artery disease.
Preclinical studies of DCB angioplasty Incremental development of technology Experience with DES technology showed that controlled drug release in the first 30–60 days after stent implantation is critical for device efficacy.8,9 Accordingly, sustained antirestenotic efficacy after a brief DCB inflation would seem to be a substantial technological challenge. Axel and colleagues provided initial evidence that single-dose exposure of vascular smooth muscle cells to paclitaxel resulted in sustained inhibition of proliferation and migration in monocultures and in co-cultures with human arterial 14 | JANUARY 2014 | VOLUME 11
endothelial cells.10 Importantly, this effect was observed at a similar magnitude when growth factors were added to the culture.10 Moreover, studies have shown that local paclitaxel delivery via microporous or perfusion balloon catheters can produce effective neointimal inhibition.11,12 Subsequently, Scheller and colleagues tested the combination of local delivery of paclitaxel with iopromide iodin ated contrast media.13 The basis of this hypothesis is the observation that iopromide adheres to the luminal surface of the coronary vasculature for a few seconds after a single injection before complete wash-out. Paclitaxel–iopromide was injected into the coronary arteries of stented pigs, and elevated tissue concentrations were detected for up to 24 h.13 Moreover, cell-culture experiments confirmed augmented inhibition of smooth muscle cell proliferation when iopromide and paclitaxel were administered together.13 Significantly lower efficacy was seen when paclitaxel alone was applied, suggesting an important role for iopromide as a carrier and solvent for paclitaxel in facilitating prolonged exposure at the vascular surface.14 The next step was to test device-facilitated local application of the iopromide–paclitaxel mixture on a coated balloon catheter. Scheller and colleagues examined the biological effect of paclitaxel DCB inflation in stented pig coronary arteries and found a significant reduction in neointimal hyperplasia in comparison with uncoated balloon catheters.15 The greatest efficacy was observed with a coating mixture of paclitaxel and acetone, supporting the concept that solvents or excipients of paclitaxel are needed to augment its vascular effect in DCB technology.15 In a subsequent study, paclitaxel DCB angioplasty was shown to be superior to sirolimus‑eluting stents at 28 days in healthy coronary pig arteries.16 On the basis of these favourable data, additional devices relying on modified or novel coating technologies were developed. Comparative effectiveness studies showed that substantial variability in reducing neointimal growth existed with different DCB catheters.17,18 In particular, devices using excipient coating showed greater inhibition of neointimal growth than DCBs with no excipient.19 Additionally, investigators in these studies documented that fibrin deposition and delayed endothelialization— known hallmarks of delayed vascular healing after DES implantation—might be important indirect signs of effective transfer of paclitaxel to the vascular wall (Figure 2a–d).18–20 Efficacy after DCB deployment is based on the rapid transfer of antiproliferative agents from the balloon surface to the vessel wall, and subsequent local tissue retention— the combination of which results in sustained tissue effects. Interestingly, in animal models, the maximum efficacy in reducing neointimal growth is seen when DCBs are combined with metallic stents, either precrimped onto the DCB balloon catheter or implanted separately immediately before DCB inflation.14,15,18–20 By contrast, deployment of DCBs followed by stent implantation resulted in less effective reduction of neointimal growth in preclinical studies.18,19 The most-likely explanation is that drug delivery into vascular tissue can be amplified by local strut-induced microinjuries. This hypothesis implies that
www.nature.com/nrcardio © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Table 1 | Currently available drug-coated balloon devices Device
Company
Coating
Drug dose (μg/mm2)
CE mark*
Danubio®
Minvasys, Paris, France
Paclitaxel–BTHC
2.5
Yes
Dior® II
Eurocor, Bonn, Germany
Paclitaxel–shellac
3.0
Yes
Elutax SV™
Aachen Resonance, Luxembourg, Luxembourg
Paclitaxel
2.0
Yes
IN.PACT™ Falcon
Medtronic Vascular, Santa Rosa, CA, USA
Paclitaxel–urea
3.0
Yes
Lutonix DCB® (Moxy)
BARD, Murray Hill, NJ, USA
Paclitaxel–polysorbate/sorbitol
2.0
No
Pantera Lux®
Biotronik, Bülach, Switzerland
Paclitaxel–BTHC
3.0
Yes
Primus™
Cardionovum, Warsaw, Poland
Paclitaxel–shellac
3.0
Yes
SeQuent® Please
B. Braun Melsungen AG, Berlin, Germany
Paclitaxel–iopromide
3.0
Yes
Advance 18 PTX™
Cook Medical, Bloomington, IN, USA
Paclitaxel
3.0
Yes
Cotavance®
Bayer Schering Pharma AG, Berlin, Germany
Paclitaxel–iopromide
3.0
Yes
Freeway™
Eurocor, Bonn, Germany
Paclitaxel–shellac
3.0
Yes
IN.PACT™ Admiral, Amphirion, Pacific
Medtronic Vascular, Santa Clara, CA, USA
Paclitaxel–urea
3.0
Yes
Lutonix DCB® (Moxy)
BARD, Murray Hill, NJ, USA
Paclitaxel–polysorbate/sorbitol
2.0
Yes
Legflow®
Cardionovum, Warsaw, Poland
Paclitaxel–shellac
3.0
Yes
Passeo‑18 Lux®
Biotronik, Bülach, Switzerland
Paclitaxel–BTHC
3.0
No
Stellarex®
Covidien, Mansfield, MA, USA
Paclitaxel
2.0
No
Coronary‡
Peripheral
§
*None of the devices is currently approved by the FDA for clinical use. ‡Coronary artery devices are compatible with 0.014 inch guidewires; balloon inflations of 30–60 s are recommended. §Peripheral artery devices are compatible with 0.014–0.035 inch guidewires; balloon inflations of up to 180 s are recommended. Abbreviation: BTHC, butyryl-tri-hexyl citrate.
moderate traumatic injury of the plaque or vessel wall is an important prerequisite for successful retention of anti restenotic drug, and that a morphologically intact vascular wall is a substantial barrier to drug penetration.21 This idea is in keeping with the recognition that thorough lesion preparation with a plain balloon, and perhaps also with a cutting or scoring balloon, before DCB angioplasty is particularly important for clinical success.22,23 Interestingly, the efficacy of DCBs with precrimped stents in preclinical trials was not corroborated in human trials applying similar sequences of vascular injury.24 A potential explanation might be the differences between healthy porcine and diseased human coronary arteries. Healthy pig coronary arteries follow a response-to-injury cascade, where neointima formation occurs rapidly as a consequence of vascular injury; however, a more-complex and delayed pattern of vascular healing is observed in diseased human coronary arteries.25 Additionally, underlying plaque morphology (such as degree of calcification) and distribution can substantially alter local tissue uptake and device efficacy.26,27
thinning with replacement of smooth muscle cells by collagen (Figure 2a–d).18–20 Therefore, models with longterm follow-up are vital elements of preclinical investigation. Moreover, improvements in coating technology might reduce vascular toxicity. An animal study published in 2013 suggested an improved safety profile with newgeneration DCB catheters using a coating process that yields more-homogenous drug distribution than with first-generation DCBs catheters.28 Examination of the remote vascular bed is another important component of preclinical investigation with DCBs,29 because embolization of DCB-coating particles into the microvascular circulation is occasionally seen (Figure 2e,f). Therefore, the microstructure of the DCB surface coating is likely to be important in limiting downstream embolization. Notably, this issue is not restricted to DCB angioplasty and has also been reported with DESs.30,31 Moreover, although microparticulate embolization might, theoretically, result in slow flow or periprocedural myocardial infarction after DCB angioplasty, these concerns have not been borne out in clinical trials to date.
Safety assessment Regarding safety assessment of DCBs, two main aspects should be considered: local vessel wall toxicity and microparticulate embolization from the balloon coating into the distal microvasculature. Vascular toxicity and delayed healing has been observed in animal studies after DCB angioplasty and is characterized by excess fibrin deposition along the vascular circumference and medial
Novel DCB technologies Currently available DCB devices share the use of paclitaxel as an antiproliferative drug, owing to its high lipophilicity and favourable tissue kinetics (Table 1). However, some interest is focused on the development of limus-based DCB catheters as a result of the favourable experience with this class of drugs in DES technology. Cremers and colleagues reported encouraging data
NATURE REVIEWS | CARDIOLOGY
VOLUME 11 | JANUARY 2014 | 15 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS a
◀ Figure 2 | Vascular effects of drug-coated balloon
b
*
c
d
e
versus plain-balloon angioplasty in a rabbit iliac model. a,b | Low-power Movat Pentachrome and high-power hematoxylin and eosin staining, respectively, after drugcoated balloon angioplasty. Note the presence of fibrin-rich thrombus formation along the luminal surface (*) with lymphohistiocytic intimal inflammation (arrow), an indirect sign of successful paclitaxel transfer to the vessel wall. c,d | Low-power Movat Pentachrome and high-power hematoxylin and eosin staining, respectively, after plainballoon angioplasty. Note the absence of fibrin and only minor inflammatory response. Note also the lifting of the endothelial layer after balloon angioplasty (arrow), a sign of balloon-induced vascular injury. e,f | Representative hematoxylin and eosin stained images of vascular changes owing to microparticulate embolization in skeletal muscle downstream from sites of drug-coated balloon angioplasty in rabbit iliac arteries at 28 days. Panel e shows luminal obstruction by fibrin (arrow) with loss of medial (M) smooth muscle cells and evidence of nuclear pyknosis/karyorrhexis. Also note the aberrant swollen nuclei (arrowheads). Panel f shows obliteration of the vessel by severe lymphohistiocytic inflammation mixed with a few red blood cells. The central region (arrow) shows foreign crystalline material, which is likely to be embolized excipient and paclitaxel matrix coating. Images courtesy of CVPath Institute, Gaithersburg, MD, USA.
f
M
with a novel zotarolimus-coated balloon.32 In another study, the drug was found to be effectively transferred from a zotarolimus-coated balloon into femoral artery tissue as early as 5 min after angioplasty, with a detectable level of the drug persisting until the 28‑day followup in a hypercholesterolaemic swine model.33 Long-term inhibition of neointimal growth remains to be tested with these devices and, because of concerns about the effectiveness of balloon-to-tissue transfer, mechanisms to sustain drug uptake might be more important with zotarolimus-coated balloons than with paclitaxel-coated balloons.34 Additionally, the use of nanoparticle excipients might be a promising approach to enhance the uptake of limus drugs.35
Trials of DCBs in coronary disease Although DCB angioplasty has theoretical advantages over stent implantation, the limitations inherent to an angioplasty-only strategy mean that clinical use has focused mainly on scenarios where stent implantation is not desirable or effective. The widest experience and mostencouraging data with DCB use in coronary intervention are in the treatment of vessels with a pre-existing stent scaffold—the setting of in-stent restenosis. Overall, data in de novo coronary disease is considerably less encouraging. 16 | JANUARY 2014 | VOLUME 11
Coronary in-stent restenosis Scheller and colleagues reported the first proof-of-concept, randomized clinical trial with a paclitaxel–iopromide DCB in 2006.36 In the landmark PACCOCATH‑ISR pilot study,36 the investigators randomly allocated 52 patients with restenosis within bare-metal stents to either DCB or plain-balloon angioplasty. The main finding was that DCB therapy significantly reduced late luminal loss at 6 months in comparison with balloon angioplasty (0.03 ± 0.48 mm versus 0.74 ± 0.86 mm; P = 0.002).36 The researchers subsequently extended enrolment in the PACCOCATH‑ISR II trial37 and reported encouraging 2‑year results with the combined dataset. The latest, 5‑year results confirm durability of DCB efficacy.38 The main limitation of these studies was the choice of control group. Technological advances meant that balloon angioplasty alone was not the benchmark comparator in the treatment of in-stent restenosis: DES therapy had already been proven to be superior to balloon angioplasty in a number of randomized, controlled trials,39,40 and meta-analysis of data had suggested that DESs should be considered the first-line treatment for patients with bare-metal in-stent restenosis.41 Therefore, any subsequent technology should have been compared with DES therapy. Accordingly, in the PEPCAD‑II trial,42 a second-generation paclitaxel–iopromide DCB (SeQuent® Please; B. Braun Melsungen AG, Berlin, Germany) was compared with a paclitaxel-eluting stent (PES; Taxus® Liberté®; Boston Scientific, Maple Grove, MN, USA). The primary end point of in-segment late loss at 6 months was lower with DCB than with PES (0.17 ± 0.42 mm versus 0.38 ± 0.61 mm; P
Introduction
Deutsches Herzzentrum, Technische Universität, Lazarettstrasse 36, D‑80636 Munich, Germany (R. A. Byrne, M. Joner, A. Kastrati). Cardiac Department, Hospital Universitario de La Princesa, IIS-IP, Universidad Autónoma de Madrid, c/ Diego de León 62, Madrid 28006, Spain (F. Alfonso). Correspondence to: A. Kastrati [email protected]
The aim of catheter-based percutaneous intervention is the durable relief of arterial obstruction with a mini mal therapeutic footprint. Balloon angioplasty and its forerunner, catheter dilatation (‘dottering’), were pioneering clinical advances. However, the principal limit ations were twofold: modest acute gain and unstable acute results (owing to arterial recoil, plaque prolapse, and vessel dissection), and reduced long-term patency (owing to constrictive remodelling and neointimal hyperplasia). Implantation of metallic or, indeed, polymeric stents results in increased acute gain and stability of acute results, and improved mechanical scaffolding. However, these improvements occur at the expense of increased vessel injury and, consequently, increased levels of neointimal hyperplasia, which is measured in studies as ‘late luminal loss’ at angiographic follow-up.1,2 Drug-eluting stents (DESs) further improve outcomes by combining high mechanical acute gain with local drug delivery to reduce late luminal loss.3,4 The collateral cost here, however, is a systematic delay in healing of the stented vessel, which is associated with a notable increase in late stent thrombosis compared with bare-metal stents, at least with early-generation DES devices.5–7 Competing interests M. Joner declares associations with the following companies: Abbott, Biotronik, Cardionovum, Medtronic, St. Jude Medical, and Zorion. A. Kastrati declares associations with the following companies: Biosensors, Biotronik, and St. Jude Medical. See the article online for full details of the relationships. R. A. Byrne and F. Alfonso declare no competing interests.
Nonstent-based local drug delivery with the use of drug-coated balloons (DCBs) is an alternative to DES technology with broad clinical applicability. The main theoretical advantages over stent-based local drug delivery include a broader area of surface contact and morehomogenous drug–tissue transfer; potential amelioration of delayed arterial healing owing to absence of long-term inflammatory nidus (no requirement for durable polymer surface coatings with DCBs); preservation or early restor ation of normal vessel anatomy and function; and preferential application in scenarios where stent implantation is not desirable (in-stent restenosis, bifurcation carina, or diffuse femoropopliteal artery disease). The limitations of DCB therapy are primarily those shared with balloon angioplasty in the early phase after intervention— namely, lower acute gain and potentially more-unstable acute results. DCB catheters are broadly similar in construction to standard angioplasty catheters (Figure 1). The main distinction is that the balloon is coated with a thin, pharmacologically active surface layer. The primary challenges from a medical engineering standpoint are effective drug transfer within the time frame imposed by a single inflation of the balloon (typically a maximum of 30–60 s in coronary intervention) and the minimization of drug loss during device handling and tracking to the lesion. The surface coating is typically composed of a mixture of fairly high-dose antiproliferative drug (most commonly paclitaxel) and a carrier substance or excipient. Currently available DCB devices and coating constituents
NATURE REVIEWS | CARDIOLOGY
VOLUME 11 | JANUARY 2014 | 13 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Key points ■■ Nonstent-based local drug delivery with drug-coated balloons (DCBs) can achieve rapid transfer of a drug to surrounding tissue, and durable antirestenotic efficacy ■■ Theoretical advantages of DCBs over drug-eluting stents (DESs) include broad surface contact, homogenous drug distribution, absence of stent footprint or polymer residue, and restoration of normal vessel anatomy and function ■■ Preclinical studies have shown that DCB delivery of paclitaxel combined with a hydrophilic spacer (excipient) results in clinically effective local tissue drug concentrations and sustained inhibition of neointimal growth ■■ Among patients with coronary bare-metal or drug-eluting stent restenosis, DCBs show similar results to standard-of-care treatment (repeat stenting with a DES) and obviate the need for further stent implants ■■ Among patients with de novo coronary artery disease, convincing data from randomized trials to support DCB therapy is lacking, and the use of DCBs for this indication requires further investigation ■■ In peripheral artery disease, DCB therapy has proven superior to balloon angioplasty for treatment of de novo femoropopliteal and below-the-knee disease
a
b
c
d
e Balloon surface Matrix coating Endothelium Active drug
Plaque or neointima Tunica media
Exipient
Figure 1 | Drug-coated balloon structure and mechanism of action in in-stent restenosis. a | Neointimal hyperplasia can slowly narrow a stented lesion, causing instent restenosis. b | The drug-eluting balloon is advanced to the lesion site. c | When the balloon is well positioned, the balloon is inflated for at least 30 s, releasing the antiproliferative drug. d | The balloon is withdrawn and the drug penetrates into the artery wall. Paclitaxel will act over several days to minimize cell regrowth. e | In most of these devices, drug release is facilitated by a coating that consists of a mixture of lipophilic drug and a hydrophilic spacer or excipient that facilitates uptake into the vessel wall. The spacer or excipient combines with the active drug to form a matrix coating that remains on the surface of the balloon catheter.
are shown in Table 1. In this Review, we discuss the use of DCBs to treat de novo or in-stent restenosis of coronary or peripheral artery disease.
Preclinical studies of DCB angioplasty Incremental development of technology Experience with DES technology showed that controlled drug release in the first 30–60 days after stent implantation is critical for device efficacy.8,9 Accordingly, sustained antirestenotic efficacy after a brief DCB inflation would seem to be a substantial technological challenge. Axel and colleagues provided initial evidence that single-dose exposure of vascular smooth muscle cells to paclitaxel resulted in sustained inhibition of proliferation and migration in monocultures and in co-cultures with human arterial 14 | JANUARY 2014 | VOLUME 11
endothelial cells.10 Importantly, this effect was observed at a similar magnitude when growth factors were added to the culture.10 Moreover, studies have shown that local paclitaxel delivery via microporous or perfusion balloon catheters can produce effective neointimal inhibition.11,12 Subsequently, Scheller and colleagues tested the combination of local delivery of paclitaxel with iopromide iodin ated contrast media.13 The basis of this hypothesis is the observation that iopromide adheres to the luminal surface of the coronary vasculature for a few seconds after a single injection before complete wash-out. Paclitaxel–iopromide was injected into the coronary arteries of stented pigs, and elevated tissue concentrations were detected for up to 24 h.13 Moreover, cell-culture experiments confirmed augmented inhibition of smooth muscle cell proliferation when iopromide and paclitaxel were administered together.13 Significantly lower efficacy was seen when paclitaxel alone was applied, suggesting an important role for iopromide as a carrier and solvent for paclitaxel in facilitating prolonged exposure at the vascular surface.14 The next step was to test device-facilitated local application of the iopromide–paclitaxel mixture on a coated balloon catheter. Scheller and colleagues examined the biological effect of paclitaxel DCB inflation in stented pig coronary arteries and found a significant reduction in neointimal hyperplasia in comparison with uncoated balloon catheters.15 The greatest efficacy was observed with a coating mixture of paclitaxel and acetone, supporting the concept that solvents or excipients of paclitaxel are needed to augment its vascular effect in DCB technology.15 In a subsequent study, paclitaxel DCB angioplasty was shown to be superior to sirolimus‑eluting stents at 28 days in healthy coronary pig arteries.16 On the basis of these favourable data, additional devices relying on modified or novel coating technologies were developed. Comparative effectiveness studies showed that substantial variability in reducing neointimal growth existed with different DCB catheters.17,18 In particular, devices using excipient coating showed greater inhibition of neointimal growth than DCBs with no excipient.19 Additionally, investigators in these studies documented that fibrin deposition and delayed endothelialization— known hallmarks of delayed vascular healing after DES implantation—might be important indirect signs of effective transfer of paclitaxel to the vascular wall (Figure 2a–d).18–20 Efficacy after DCB deployment is based on the rapid transfer of antiproliferative agents from the balloon surface to the vessel wall, and subsequent local tissue retention— the combination of which results in sustained tissue effects. Interestingly, in animal models, the maximum efficacy in reducing neointimal growth is seen when DCBs are combined with metallic stents, either precrimped onto the DCB balloon catheter or implanted separately immediately before DCB inflation.14,15,18–20 By contrast, deployment of DCBs followed by stent implantation resulted in less effective reduction of neointimal growth in preclinical studies.18,19 The most-likely explanation is that drug delivery into vascular tissue can be amplified by local strut-induced microinjuries. This hypothesis implies that
www.nature.com/nrcardio © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS Table 1 | Currently available drug-coated balloon devices Device
Company
Coating
Drug dose (μg/mm2)
CE mark*
Danubio®
Minvasys, Paris, France
Paclitaxel–BTHC
2.5
Yes
Dior® II
Eurocor, Bonn, Germany
Paclitaxel–shellac
3.0
Yes
Elutax SV™
Aachen Resonance, Luxembourg, Luxembourg
Paclitaxel
2.0
Yes
IN.PACT™ Falcon
Medtronic Vascular, Santa Rosa, CA, USA
Paclitaxel–urea
3.0
Yes
Lutonix DCB® (Moxy)
BARD, Murray Hill, NJ, USA
Paclitaxel–polysorbate/sorbitol
2.0
No
Pantera Lux®
Biotronik, Bülach, Switzerland
Paclitaxel–BTHC
3.0
Yes
Primus™
Cardionovum, Warsaw, Poland
Paclitaxel–shellac
3.0
Yes
SeQuent® Please
B. Braun Melsungen AG, Berlin, Germany
Paclitaxel–iopromide
3.0
Yes
Advance 18 PTX™
Cook Medical, Bloomington, IN, USA
Paclitaxel
3.0
Yes
Cotavance®
Bayer Schering Pharma AG, Berlin, Germany
Paclitaxel–iopromide
3.0
Yes
Freeway™
Eurocor, Bonn, Germany
Paclitaxel–shellac
3.0
Yes
IN.PACT™ Admiral, Amphirion, Pacific
Medtronic Vascular, Santa Clara, CA, USA
Paclitaxel–urea
3.0
Yes
Lutonix DCB® (Moxy)
BARD, Murray Hill, NJ, USA
Paclitaxel–polysorbate/sorbitol
2.0
Yes
Legflow®
Cardionovum, Warsaw, Poland
Paclitaxel–shellac
3.0
Yes
Passeo‑18 Lux®
Biotronik, Bülach, Switzerland
Paclitaxel–BTHC
3.0
No
Stellarex®
Covidien, Mansfield, MA, USA
Paclitaxel
2.0
No
Coronary‡
Peripheral
§
*None of the devices is currently approved by the FDA for clinical use. ‡Coronary artery devices are compatible with 0.014 inch guidewires; balloon inflations of 30–60 s are recommended. §Peripheral artery devices are compatible with 0.014–0.035 inch guidewires; balloon inflations of up to 180 s are recommended. Abbreviation: BTHC, butyryl-tri-hexyl citrate.
moderate traumatic injury of the plaque or vessel wall is an important prerequisite for successful retention of anti restenotic drug, and that a morphologically intact vascular wall is a substantial barrier to drug penetration.21 This idea is in keeping with the recognition that thorough lesion preparation with a plain balloon, and perhaps also with a cutting or scoring balloon, before DCB angioplasty is particularly important for clinical success.22,23 Interestingly, the efficacy of DCBs with precrimped stents in preclinical trials was not corroborated in human trials applying similar sequences of vascular injury.24 A potential explanation might be the differences between healthy porcine and diseased human coronary arteries. Healthy pig coronary arteries follow a response-to-injury cascade, where neointima formation occurs rapidly as a consequence of vascular injury; however, a more-complex and delayed pattern of vascular healing is observed in diseased human coronary arteries.25 Additionally, underlying plaque morphology (such as degree of calcification) and distribution can substantially alter local tissue uptake and device efficacy.26,27
thinning with replacement of smooth muscle cells by collagen (Figure 2a–d).18–20 Therefore, models with longterm follow-up are vital elements of preclinical investigation. Moreover, improvements in coating technology might reduce vascular toxicity. An animal study published in 2013 suggested an improved safety profile with newgeneration DCB catheters using a coating process that yields more-homogenous drug distribution than with first-generation DCBs catheters.28 Examination of the remote vascular bed is another important component of preclinical investigation with DCBs,29 because embolization of DCB-coating particles into the microvascular circulation is occasionally seen (Figure 2e,f). Therefore, the microstructure of the DCB surface coating is likely to be important in limiting downstream embolization. Notably, this issue is not restricted to DCB angioplasty and has also been reported with DESs.30,31 Moreover, although microparticulate embolization might, theoretically, result in slow flow or periprocedural myocardial infarction after DCB angioplasty, these concerns have not been borne out in clinical trials to date.
Safety assessment Regarding safety assessment of DCBs, two main aspects should be considered: local vessel wall toxicity and microparticulate embolization from the balloon coating into the distal microvasculature. Vascular toxicity and delayed healing has been observed in animal studies after DCB angioplasty and is characterized by excess fibrin deposition along the vascular circumference and medial
Novel DCB technologies Currently available DCB devices share the use of paclitaxel as an antiproliferative drug, owing to its high lipophilicity and favourable tissue kinetics (Table 1). However, some interest is focused on the development of limus-based DCB catheters as a result of the favourable experience with this class of drugs in DES technology. Cremers and colleagues reported encouraging data
NATURE REVIEWS | CARDIOLOGY
VOLUME 11 | JANUARY 2014 | 15 © 2014 Macmillan Publishers Limited. All rights reserved
REVIEWS a
◀ Figure 2 | Vascular effects of drug-coated balloon
b
*
c
d
e
versus plain-balloon angioplasty in a rabbit iliac model. a,b | Low-power Movat Pentachrome and high-power hematoxylin and eosin staining, respectively, after drugcoated balloon angioplasty. Note the presence of fibrin-rich thrombus formation along the luminal surface (*) with lymphohistiocytic intimal inflammation (arrow), an indirect sign of successful paclitaxel transfer to the vessel wall. c,d | Low-power Movat Pentachrome and high-power hematoxylin and eosin staining, respectively, after plainballoon angioplasty. Note the absence of fibrin and only minor inflammatory response. Note also the lifting of the endothelial layer after balloon angioplasty (arrow), a sign of balloon-induced vascular injury. e,f | Representative hematoxylin and eosin stained images of vascular changes owing to microparticulate embolization in skeletal muscle downstream from sites of drug-coated balloon angioplasty in rabbit iliac arteries at 28 days. Panel e shows luminal obstruction by fibrin (arrow) with loss of medial (M) smooth muscle cells and evidence of nuclear pyknosis/karyorrhexis. Also note the aberrant swollen nuclei (arrowheads). Panel f shows obliteration of the vessel by severe lymphohistiocytic inflammation mixed with a few red blood cells. The central region (arrow) shows foreign crystalline material, which is likely to be embolized excipient and paclitaxel matrix coating. Images courtesy of CVPath Institute, Gaithersburg, MD, USA.
f
M
with a novel zotarolimus-coated balloon.32 In another study, the drug was found to be effectively transferred from a zotarolimus-coated balloon into femoral artery tissue as early as 5 min after angioplasty, with a detectable level of the drug persisting until the 28‑day followup in a hypercholesterolaemic swine model.33 Long-term inhibition of neointimal growth remains to be tested with these devices and, because of concerns about the effectiveness of balloon-to-tissue transfer, mechanisms to sustain drug uptake might be more important with zotarolimus-coated balloons than with paclitaxel-coated balloons.34 Additionally, the use of nanoparticle excipients might be a promising approach to enhance the uptake of limus drugs.35
Trials of DCBs in coronary disease Although DCB angioplasty has theoretical advantages over stent implantation, the limitations inherent to an angioplasty-only strategy mean that clinical use has focused mainly on scenarios where stent implantation is not desirable or effective. The widest experience and mostencouraging data with DCB use in coronary intervention are in the treatment of vessels with a pre-existing stent scaffold—the setting of in-stent restenosis. Overall, data in de novo coronary disease is considerably less encouraging. 16 | JANUARY 2014 | VOLUME 11
Coronary in-stent restenosis Scheller and colleagues reported the first proof-of-concept, randomized clinical trial with a paclitaxel–iopromide DCB in 2006.36 In the landmark PACCOCATH‑ISR pilot study,36 the investigators randomly allocated 52 patients with restenosis within bare-metal stents to either DCB or plain-balloon angioplasty. The main finding was that DCB therapy significantly reduced late luminal loss at 6 months in comparison with balloon angioplasty (0.03 ± 0.48 mm versus 0.74 ± 0.86 mm; P = 0.002).36 The researchers subsequently extended enrolment in the PACCOCATH‑ISR II trial37 and reported encouraging 2‑year results with the combined dataset. The latest, 5‑year results confirm durability of DCB efficacy.38 The main limitation of these studies was the choice of control group. Technological advances meant that balloon angioplasty alone was not the benchmark comparator in the treatment of in-stent restenosis: DES therapy had already been proven to be superior to balloon angioplasty in a number of randomized, controlled trials,39,40 and meta-analysis of data had suggested that DESs should be considered the first-line treatment for patients with bare-metal in-stent restenosis.41 Therefore, any subsequent technology should have been compared with DES therapy. Accordingly, in the PEPCAD‑II trial,42 a second-generation paclitaxel–iopromide DCB (SeQuent® Please; B. Braun Melsungen AG, Berlin, Germany) was compared with a paclitaxel-eluting stent (PES; Taxus® Liberté®; Boston Scientific, Maple Grove, MN, USA). The primary end point of in-segment late loss at 6 months was lower with DCB than with PES (0.17 ± 0.42 mm versus 0.38 ± 0.61 mm; P
![[Peripheral artery disease and acute coronary syndrome].](/assets/img/hold.png)
