Coronary bifurcation lesion (CBL) is a common finding in daily practice, being encountered in up to 20% of percutaneous coronary interventions (PCI) [1]. Despite the common occurrence, CBL treatment still remains a challenging scenario because of its technical complexity and the not always favorable clinical outcomes. In fact, coronary bifurcation is a complex anatomical structure composed of three different vessel segments: proximal main vessel (MV), distal MV and side branch (SB). Adequate angiographic result in all segments has to be pursued, even if the definition of optimal SB result has yet to be established [2]. Drug-coated balloons (DCBs) are modern devices able to guarantee a fast and homogenous transfer of anti-proliferative drugs into the vessel, without the permanent implant of metallic struts. This “leaving nothing behind” philosophy re-proposes some of the advantages related to the bioresorbable scaffold use [3]. On this premise, in-stent restenosis, small vessel and diffuse de-novo coronary artery disease have been, so far, considered the ideal scenario for DCBs application.

The European Bifurcation Club (EBC) has recently recommended that PCI on bifurcation stenting should adhere to the “keep it simple and safe” principle, trying to limit the number of stents [2]. On this basis, a one-stent strategy is usually the preferred approach for the vast majority of CBLs (the so-called “provisional strategy”), whereas a two-stent approach should be selected in patients with complex lesions involving large and diseased SBs (especially unprotected left main). These recommendations go hand in hand with the “leaving nothing behind” philosophy, making the DCB very attractive in the CBL setting for several reasons. First of all, DCB can dramatically increase the rate of provisional strategy, reducing the incidence of device-related failure (in-stent restenosis and stent thrombosis) associated with a wide application of two-stent strategies. Moreover, DCB can reduce the PCI complexity, since two-stent strategies require extensive knowledge of technical steps. Secondly, treating diseased SBs with anti-proliferative drug can provide better results in comparison with a conventional dilation, mitigating the limits of the current SB plain angioplasty recommended among provisional strategy steps. Lastly, CBL SBs are often small vessels (with a diameter $\le$2.75 mm) but subtending a not negligible area of myocardium. Taken together, these aspects represent the rationale of different studies investigating the efficacy and safety of DCB application in CBL setting. In this review, we provide a contemporary overview on the topic, discussing the available evidence; articles were searched on the online library Pubmed.gov on the basis of the following keywords: “drug-coated balloon; drug-eluting balloon; coronary bifurcation lesions”. All the articles including CBLs patients treated with DCB were included.

In several trials, DCBs have been proved to be a safe and effective therapeutic option in the context of de-novo coronary artery disease [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27]. Despite hundreds of patients have been enrolled in these trials, no definite conclusions can be formulated concerning the role of DCB in CBL setting. In fact, in most of the above-named trials the percentage and the number of CBL patients have not been reported [5, 6, 8, 10, 11, 13, 14, 15, 22, 24, 27] or CBL has represented an exclusion criteria [7, 12, 16, 17, 21]. Thus, the remaining of the afore mentioned studies enrolled a small number of CBL patients [4, 9, 18, 19, 20, 23, 25], suggesting conflicting results (Table 1, Ref. [4, 9, 18, 19, 20, 23, 25]): in the DEBUT trial [23], DCBs have been associated with a better outcome (composite of cardiovascular mortality, nonfatal myocardial infarction, or ischemia-driven target lesion revascularization at 9 months) in comparison with bare-metal stent (BMS), whereas in the BASKET-SMALL 2 trial [20] no difference has been emerged between DCB and drug-eluting stent (DES) at the 1-year assessment. Recently, Iannopollo et al. [26], described a multicenter registry with the main aim to assess the performance of the Agent DCB (Boston Scientific) in all PCI settings, including 97 patients with 117 CBLs treated. In this CBL subgroup, the treated bifurcations were mainly true bifurcations (84%) and the adopted stent strategy was provisional, with DES implantation in the MV and DCB in the SB. Authors reported optimal procedural outcomes (92% of success defined as completion of the procedure with no in-lab complications, final Thrombolysis in Myocardial Infarction flow 3, and residual stenosis 30%) and favorable 1-year clinical outcomes (3.7% rate of major adverse cardiac events, all of them consisting in target-lesion revascularization [TLR]).

Considering the small number of CBL patients enrolled in the main trials assessing the use of DCBs in the context of de-novo coronary artery disease, the need of drawing focused studies was primary. On this premise, several studies have been so far conducted. However, the available studies, focused on the topic, are characterized by heterogeneous designs, especially considering the type of bifurcation studied (according to Medina classification), the applied strategy and the step in which DCB was delivered. Not all bifurcations are equal in complexity, being classifiable in three major types: CBLs involving only the SB (Medina 0.0.1), CBLs involving only the MV (Medina 1.0.0 or 0.1.0 or 1.1.0) and CBLs involving both MV and SB (true bifurcation; Medina 1.0.1 or 0.1.1 or 1.1.1). Briefly, two strategies are possible: (1) MV stenting plus DCB application (in SB only or in both segments); (2) DCB-only (both in SB and MV). Different combinations of these approaches have been described in the available trials. Yet, also the timing of DCB application can introduce variability. In fact, when a DCB is used to treat SBs can be inflated before MV stenting (then avoiding SB rewiring and final kissing-balloon inflation), during the kissing-balloon inflation (after MV stenting) and even after kissing-balloon inflation (as final SB dilation). Different theoretical advantages and disadvantages can influence the result: pre-dilating the SB with DCB application may results in SB dissection; on the other hand, applying DCB after MB stenting could result in partial disruption of the coating caused by the stent strut, taking into account the profile of such balloons. Despite the best timing is still unknown, some Authors suggest that DCB should be applied as the final step and not as an intermediate one [28]. Lastly, a wide variety of different DCBs have been used in the different studies: interaction among doses, formulations, and release kinetics of the drugs used play an important role in determining no evidence of a “class effect” among different platforms [29].

On these premises, an overview of the available studies is reported in Table 2 (Ref. [30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42]).

Despite the first studies [30, 31, 32] documented the safety and the efficacy of the strategy consisting in DCB application followed by MV stenting, their main limitation consisted in the use of BMS, nowadays considered an outdated tool. In fact, the randomized controlled trials (RCTs) comparing DCB inflation plus BMS with standard PCI performed with BMS or DES, showed the absence of angiographic and clinical superiority of the first strategy over conventional BMS and the inferiority compared with DES [33, 35].

Three observational studies [36, 37, 38] assessed the feasibility of the hybrid approach (DES implantation in MV and DCB inflation in SB). In the multicenter BIOLUX-trial [36] 35 CBL patients deemed appropriate for provisional stenting technique were treated with DCB (Pantera Lux, Biotronik AG, Buelach, Switzerland) inflation in SB before DES implantation in MV. Authors reported a 60% device success (defined as procedural absence of residual SB diameter stenosis $\ge$30%), a low 9-month late lumen loss (LLL) and a 5.9% rate of 12-month major adverse cardiovascular events (MACE). Despite the mandatory final kissing-balloon inflation which represents a strength of this study, major limitation consists in the exclusion of major anatomical bifurcations (LM and very proximal CBLs). The French DEBSIDE trial [36] treated 50 CBL patients with DCB inflation (DANUBIO, Minvasys, Gennevilliers, France) in SB, after the systematic implantation of DES (Nile PAX, Minvasys, Gennevilliers, France) in MV. Procedural success was 100%. 6-month ostial SB LLL was the pre-specified primary endpoint and resulted in a very low value (–0.04 $\pm$ 0.34 mm). One of the main limitations of this study is the monobrand nature of the used devices, that are not anymore available on the market. The third observational study, named SARPEDON trial [38], enrolled 58 CBL patients in which a final DCB (Pantera Lux, Biotronik AG, Buelach, Switzerland) application in SB followed MV stenting. The 6-month angiographic follow-up showed a low rate of SB LLL with a total 10% rate of restenosis (6% in the MV and 4% in the SB), whereas MACE occurred in 19% of the patients at 1-year follow-up. Taken together, despite the absence of a control group and the limited number of enrolled patients, the results provided by the afore mentioned three trials showed a good performance of the hybrid strategy.

In the last years, the concept to treat de-novo CBLs only with DCB avoiding stent implantation (DCB-only strategy) has quickly developed. To confirm this trend, the last evidence available in the literature has been focused on DCB-only strategy [39, 40, 41, 42]. In the PEPCAD-BIF trial [39] 64 CBL patients were randomized to plain angioplasty (POBA) or DCB-only strategy. Left main bifurcation as well all bifurcations with proximal MV involvement (Medina 1.X.X) were excluded. The main study finding was a lower 9-month LLL and binary restenosis incidence in the DCB subgroup when compared with POBA subgroup. The main limitation of this study is intuitive and consists of the outdated strategy adopted for the control group (POBA). To overcome this drawback, Bruch et al. [40] compared a DCB-only strategy to a DCB plus stenting strategy, showing no differences in terms of MACE and TLR after 9 months. Nevertheless, once again, the control study group does not represent the contemporary standard-of-care, since the MV stenting was performed using BMS. In 2016, Vaquerizo et al. [41] applied a DCB-only strategy for 49 patients affected by Medina 0.0.1 CBLs. After a mandatory SB lesion preparation (as confirmed by a 59% use of cutting balloon), paclitaxel DCB (Dior, Eurocor GmbH, Bonn, Germany) was inflated for a minimum of 45 seconds. Angiographic success was 86% (in the remaining 14% a BMS was implanted), 7-month angiographic follow-up revealed 22.5% of binary restenosis whereas at 1-year follow-up MACE were 14.3%. Recently, Kitani S et al. [42] enrolled in the DCA/DCB Registry 129 patients affected by major CBLs in which DCB application in the SB was performed after lesion preparation with directional atherectomy: authors reported a low rate of TLR (3.1%) at 1-year follow-up.

Taken these data together, two meta-analyses concluded that DCB is superior to plain angioplasty for SB treatment in bifurcations [43, 44]. In the first one [43], 349 patients were included: the angiographic follow-up performed after a mean of 9 months demonstrated that DCB was associated with a lower SB LLL compared to POBA [mean difference, –0.19 mm; 95% confidence interval (CI), –0.37 to –0.01; p = 0.04] but without difference in terms of risk of SB binary restenosis [odds ratio (OR), 0.52; 95% CI, 0.18 to 1.47; p = 0.22] and 15-month MACE (OR, 0.76; 95% CI, 0.4 to 1.4; p = 0.40) and TLR (OR, 0.85; 95% CI, 0.3 to 2.4; p = 0.76). The second meta-analysis [44] is focused on comparison of DCB versus POBA for SB-only treatment in terms of LLL: including 281 CBLs, Authors showed a statistically significant difference favoring DCB over POBA [mean difference –0.24 mm; 95% CI, –0.44 to –0.05; p = 0.01].

To date, no robust data coming from RCTs and comparing DCB-only strategy versus contemporary gold-standard approach (DES) are available.

Despite DCB use in the SB is an attractive approach, according to the last EBC consensus document [3], many questions, including the appropriate SB selection, technique (DCB with or without final kissing ballooning or repeat proximal optimization technique) and actual impact on meaningful clinical endpoints are still unanswered. Studies, so far available, exploring DCB efficacy in de novo CBL had major limitations and to date have provided no conclusive evidence. Contrariwise, DCB use in bifurcation restenosis (especially after 2-stent techniques) is highly encouraged: the strategy proved to be feasible and able to minimize metal within the bifurcation [45].

Whenever a DCB strategy is chosen, a meticulous lesions preparation is suggested. The choice of DCB size can be difficult for MV treatment because of caliper change: in such case, using the distal MV reference diameter is reasonable when the distal-to-proximal MV ratio is in the range of 0.75:1 to 1:1 [29]. In the case of a DCB-only strategy, all the steps can be performed in a sequential manner, reserving kissing balloon inflation (with DCBs) in selected cases.

In our opinion, a case by case evaluation is strongly recommended: DCB can represent a valuable tool to treat SBs, especially when the amount of the disease is limited in extension (e.g., ostial stenosis), the bifurcation have disfavoring anatomical characteristics (e.g., prohibitive angle and relevant difference in branches diameters), and the subtended myocardium is relatively limited. Application of a complete DCB-only strategy (for both SB and MV) needs further investigations.

Nowadays, the best treatment of bifurcations has not yet been defined. The available studies included heterogeneous populations and specific subgroups (diabetic patients, major CBLs) have been frequently excluded. A clear definition of procedural steps in case of DCB application should be explored. RCTs comparing a strategy of DCB application versus the gold-standard treatment (DES) for SB have not been conducted. Furthermore, the optimal antiplatelet regimen and duration after DCB application are unknown. Future studies are called to answer to these unresolved questions: results are awaited from the Hyper Pilot study [46], in which a hybrid approach (DCB plus DES) has been applied to treat diffuse de-novo coronary artery disease, including bifurcation lesion (with reference diameter of the SB $\ge$2 mm and 2.75 mm). However, to the best of our knowledge no specific RCTs focused on CBLs treatment comparing DCB and DES are in pipeline, making the ongoing gaps of knowledge a relevant issue.

The “leaving nothing behind” philosophy associated with the use of DCB is an attractive therapeutic option for CBLs. Limited data have demonstrated the safety and the feasibility of this approach. However, larger RCTs with homogeneous populations and procedural steps, comparing DCB strategies to the gold standard treatment (DES implantation) are strongly needed.

Conceptualization—AI, AB, DM, MP, GDB, MT; Writing - original draft preparation—AI, AB, DM, MP, GDB, MT; Writing - review and editing—AI, AB, DM, MP, GDB, MT; Supervision—AI.

Not applicable.

We would like to express our gratitude to all those who helped us during the writing of this manuscript. Thanks to all the peer reviewers for their opinions and suggestions.

This research received no external funding.

The authors declare no conflict of interest.