Four-Year Outcomes of Left Main Percutaneous Coronary Intervention with a Bioresorbable Scaffold in the Circumflex Ostium

Objectives The study aimed to investigate the long-term outcomes of a double stent scaffold strategy in patients with left main (LM) bifurcation lesions involving the ostium of the left circumflex artery (LCX), utilizing a drug-eluting stent (DES) in the LM extending into the left anterior descending artery (LAD) and a bioresorbable vascular scaffold (BVS) in the LCX ostium. Background The high occurrence of in-stent restenosis of the LCX ostium is the major limitation of percutaneous coronary intervention (PCI) for LM lesions with a two-stent strategy. Methods This was a single-center, prospective, single-arm study of 46 consecutively enrolled patients with a stable coronary artery disease and significant unprotected LM distal bifurcation disease. Patients underwent imaging-guided PCI using DES in the LM-LAD and BVS in the LCX using a T-stent or mini-crush technique. The primary outcome at four years was the composite of death, myocardial infarction, stroke, and target lesion revascularization (TLR). Results At four years, the primary outcome was identified in 9 patients (19.6%). All events were TLRs except one myocardial infarction due to BVS thrombosis. Seven of the eight TLRs were a result of side branch BVS restenosis. Univariate predictors of the 4-year outcome were higher LDL cholesterol and BVS size ≤2.5 mm. On multivariate analysis, LCX lesion preparation with a cutting balloon and post-procedure use of intravascular ultrasound for optimization were found to be independent protective factors of MACE. Conclusions In selected patients with LM distal bifurcation disease, an imaging-guided double stent scaffold strategy with DES in the LM and BVS in the LCX ostium was technically successful in all patients and was reasonably safe and effective for four years.


Introduction
Te popularity of percutaneous coronary intervention (PCI) for the treatment of unprotected left main (LM) lesions has increased in the past few years because of advances in stent technologies and adjunctive pharmacotherapies. However, PCI of complex bifurcation lesions is still challenging in the drug-eluting stent (DES) era, and results are worse compared with simple coronary lesions. Although the provisional stent (PS) strategy is generally considered the default strategy for bifurcation lesions; there are scenarios in which the 2-stent strategy is initially necessary to improve the patency of both the main branch and the side branch. Unfortunately, the high occurrence of in-stent restenosis (ISR) of the left circumfex artery (LCX) ostium is a major limitation of the 2-stent strategy for unprotected left main (ULM) lesion [1], with a recent study reporting an ISR rate of 25.4% with this strategy (majority in the LCX ostium), vs. a rate of 6.3% with a provisional stent [2]. Conversely, the 3year outcomes of the DKCRUSH-V Trial demonstrated that target lesion failure (TLF) was higher in the PS group vs. the DK crush group (16.9% vs. 8.3% p � 0.005) and mainly driven by increased target vessel myocardial infarction (5.8% vs. 1.7%; p � 0.017) and target lesion revascularization (10.3% vs. 5.0%; p � 0.029). Te defnite or probable stent thrombosis (ST) rate at three years was 4.1% in the PS group and 0.4% in the DK group (p � 0.006). Notably, DK crush was associated with a signifcant reduction in both primary and secondary endpoints for patients with complex lesions or at high risk [3].
BVS has the unique ability to restore vascular physiology and anatomical integrity, such as native tortuosity and angulation, with only a temporary scafold necessary to maintain the patency of the vessel after the intervention [4,5]. Studies have shown complete resorption of scafold struts at 5 years [6,7]. Terefore, it may provide a novel way to treat unprotected left main distal bifurcation lesions that would beneft from a two-stent strategy at the time of intervention but leave nothing behind to preclude later surgical revascularization or noninvasive imaging.
In the present study, we sought to assess the safety and efcacy at four years post-procedure of an IVUS-guided and OCT-optimized two-stent technique (mini-crush or T-stent strategy) using an everolimus-eluting platinum chromium coronary stent with a bioresorbable polymer coating (Synergy, Boston Scientifc, Marlborough, MA) in the LM to left anterior descending (LAD) artery and a bioresorbable vascular scafold (ABSORB, Abbott Vascular, Redwood City, CA) in the LCX ostium for the treatment of distal ULMCA true bifurcation stenosis.

Study Population.
Tis was a prospective single-arm, single-center study of 46 consecutively patients enrolled from November, 2012, to December, 2015. Te study included patients with a stable coronary artery disease and angiographically signifcant unprotected LM/LAD disease in whom revascularization with a two-stent technique was preplanned. Te study inclusion criteria included patients >18 years old with symptomatic unprotected LM disease, including angiographic evidence of >50% diameter stenosis of a de novo true LM bifurcation lesion (Medina 1, 1, 1 or 0, 1, 1). Clinical exclusion criteria were acute myocardial infarction or stroke, anemia (Hb < 9 g/dl), and suspected intolerance to 1 of the study drugs. Clinical and angiographic (IVUS and OCT recommended) follow-up was conducted at 1 and 4 years post-procedure. Te study included three patients (6.5%) with high SYNTAX scores (>32) who refused surgery and had been judged to be suitable candidates for PCI. Patients that did not return to the clinic were contacted for telephone follow-up. Te protocol was approved by the ethics committee (ethics committee for clinical research at development society of Pauls Stradins Clinical University Hospital), and written consent was obtained from all patients. Te study was funded in part by the national research program "biomedicine for public health" (BIOMEDICINE) VPP 2014/VPP2014-2017.

Procedures.
All enrolled patients were scheduled to have LM PCI with two stents. A maximum of 24 hours before the procedure, all patients received 100 mg of aspirin. A loading dose of 300 mg of clopidogrel or ticagrelor 180 mg was administered before the index procedure. After the intervention, all patients received 100 mg/day aspirin for life and clopidogrel (75 mg/day) or ticagrelor 90 mg × 2 for at least 12 months. Unless precluded due to anatomical or clinical considerations, the treatment strategy was as follows: Lesion assessment was performed on the main and side branches by angiography in two opposite projections and with intravascular ultrasound (iMAP, Boston Scientifc, Marlborough, MA).
Modifcation of the atherosclerotic plaque was then performed with a cutting balloon (CB). Te CB intervention was performed with a balloon-to-vessel ratio of 1 : 1, according to IVUS media-to-media to the vessel at the lesion site, and covered the entire lesion length. Balloon infations were performed 3 times with increasing pressure throughout the lesion [8]. Standard PCI techniques were used to deploy a metallic DES in the LM/LAD and a BVS in the LCX ostium. Te choice of a T-stent or mini-crush technique was at the discretion of the operator. Postdilatation with a noncompliant balloon and a fnal kissing balloon dilatation was then performed, followed by fnal intracoronary imaging with OCT (Lightlab Imaging, Westford, MA) and 40 MHz-IVUS (Boston Scientifc Corporation, Marlborough, MA).
Procedural success was defned as a fnal residual stenosis <30% with TIMI fow grade 3 in either the main branch or the side branch.
Te study fowchart is presented in Figure 1.

Outcomes.
Te primary outcome measure was a composite of death, MI, stroke, and TLR, at 4 years. MI was diagnosed if the plasma level of the creatine kinase-myocardial band and/or troponin I/T increased to more than 3 times the upper limit of normal in no fewer than 2 blood samples. All deaths were considered cardiac in origin unless noncardiac reasons were indicated. TLR was defned as any repeat revascularization (percutaneous coronary intervention or CABG) for target lesions, respectively, in the presence of symptoms or objective signs of ischemia. ST was defned according to the Academic Research Consortium defnition. All patients were scheduled for clinical and angiographic (recommended IVUS and OCT) follow-up at 1 year and 4 years. If the patient could not visit the hospital, clinical follow-up was obtained by telephone contact. In addition to the protocol-mandated follow-up, repeat coronary angiography was performed if clinically indicated due to symptoms and signs of myocardial ischemia.

Quantitative Coronary Analysis, IVUS, and OCT.
All angiographic (pre, post, and follow-up) characteristics were analyzed ofine by quantitative coronary angiography (QCA) using Cardiovascular Angiography Analysis System (CAAS) software 5.11.2, 2013 (Pie Medical Imaging, Maastricht, Te Netherlands). QCA data included reference diameter and minimal luminal diameter at 6 lesion locations (ostial and body left main, proximal and mid part LAD, ostial and mid part LCX) where the landing of stent and scafold was planned. Preprocedural and postprocedural intravascular ultrasound (IVUS) and optical coherence tomography (OCT) at the index procedure and follow-up were used according to our center protocol to obtain more accurate information on the disease status and evaluate stent apposition after postdilatation.

Statistical Analysis.
All analyses were performed from the time of enrollment in the intention to treat population. Data were analyzed using SPSS version 21.0 (Statistical Package for the Social Sciences) software. Continuous variables are expressed as the mean ± SD. Categoric variables were tested using contingency table analyses (exact or Chi-square approximations), and continuous variables were tested using an unpaired student's t-test. Univariate logistic regression was used to identify the predictors of the primary outcome. Variables with probability values < 0.05 in individual analyses were included in the multivariate analysis. A p value of <0.05 was considered to be statistically signifcant.

Results
Patient baseline demographic and clinical characteristics are presented in Table 1. Te mean age of patients was 66.1 ± 8.0 years, and the majority were males (73.9%). Dominant cardiovascular risk factors were arterial hypertension (95.7%) and dyslipidemia (87.0%). Te history of myocardial infarction was present in 30.4% of patients; 45.7% prior had a percutaneous coronary intervention, 13.0% with non-insulin-dependent diabetes mellitus, family history of cardiovascular disease in 26.1%, and 63% of patients had chronic heart failure with a mean residual ejection fraction of 57.1 ± 9.2%. Tere are no data on statin therapy before the study, but the mean LDL at the time of the index procedure was 2.2 ± 0.8 mmol/l.

Procedural Characteristics and Quantitative Coronary
Angiography. Procedural characteristics are presented in Table 2 and QCA data are presented in Table 3. Of the 46 procedures, 65.2% were performed through the femoral approach. Prior to predilation, IVUS of both branches was performed in 43 patients (93.5) %, CB predilation in LM-      Table 4.

Predictors of Primary Outcome Measures.
Using univariate logistic regression analysis, we identifed parameters that predict primary outcome events. Results are presented in Table 5. Age, gender, and syntax score were not statistically signifcant predictors of events at 4 years, but LDL cholesterol was (OR 3.91; 95% CI 1.40-11.00). Procedural parameters that were signifcant predictors of an event were side branch plaque modifcation with cutting balloon, absorb scafold diameter ≤2.5 mm in the side branch, and IVUS use post-procedure.

. Discussion
To our knowledge, this is the frst reported series on the treatment of true left main coronary artery bifurcation with a novel hybrid stent/scafold strategy using a DES in the main branch and a BVS in the side branch. Considering the complexity of the patient group, this hybrid 2 stent/scafold strategy demonstrated acceptable efcacy and safety with a rate of major adverse events similar to that reported in other studies of provisional and dual stent treatment strategies [3,[9][10][11][12][13]. In our study, 4-year mortality after PCI with a stent/ scafold hybrid approach for true left main bifurcation lesions was 0.0%. Te NOBLE and EXCEL studies showed a higher mortality rate at 5 years in both the PCI (NOBLE 9%, EXCEL 13.0%) and CABG groups (NOBLE 9%, EXCEL 9.9%) [12,13]. In the EXCEL study, the left main stenosis involved bifurcation in 81% of patients. A planned two-stent strategy was used in 35.0% of these patients. Te death rate at 3 years was reported to be 11.0% [14]. Te DKCRUSH-V trial was a randomized comparison of the double kissing (DK) crush planned 2-stent technique and provisional 1-stent technique in distal left main bifurcation lesions [15]. In the DK crush stenting group, the mortality rate at 3 years was 6.7% [3]. In a recently published European bifurcation club left main coronary stent study (EBC MAIN), in a randomized comparison of stepwise provisional vs. systematic double stenting strategies, the death rate in the dual stent group at 1 year was 4.2% [16]. Te lower mortality in our study could be explained by the use of intracoronary imaging-guided pre-dilatation, stent sizing, and post-dilatation, or so-called iPSP strategy, which have been associated with a lower risk of cardiac events, including cardiac death, among patients undergoing PCI in complex coronary lesions [17]. In our study, preintervention and post-intervention imaging were used in more than 90% of our patients, while in EBC, main and DKCRUSH-V intravascular assessment was done in approximately 40% of the cases [15,16]. Before stent implantation, plaque modifcation of the side branch and the main branch was performed in about 80% of the cases. Among patients randomized to a planned two-stent strategy in the EBC Main, plaque pretreatment with cutting balloon, rotablation, or lithotripsy was undertaken in 22% of main vessels and 15% of side branches. Te lower mortality rate in the present study may be in part related to our learning curve gained from the frst randomized trial (2004)(2005)(2006) comparing DES and bare-metal stent for left main intervention, which showed a 2% death rate at 6 months and 7% at 3 years [8,18].
In our study, 19.6% of patients underwent target lesion revascularization (TLR) 4 years after PCI. Tis fnding is similar to the EXCEL study, where patients with planned two-stent strategies had 16.3% ischemia-driven revascularization of the left main complex at 3 years [14]. However, a lower TLR rate (5.0% at 3 years) was observed in the DK crush stenting group of the DKCRUSH-V trial [3]. Several factors could explain the diferences in the TLR rate. Gaining experience with a new technique requires time and training. It should be noted that the majority of TLR cases (six in nine) occurred during the frst 14 months of enrollment. Followup angiography could be associated with an increased incidence of TLR. It should be noted that 71.7% of our patients underwent angiographic follow-up at 4 years. In comparison, in the EXCEL study, only clinical follow-up was performed, but in the DKCRUSH-V trial, angiographic followup at 13 months and clinical follow-up at 3 years were performed. In the EBC MAIN study, the TLR rate at 1 year was 9.3% in the dual stent group [16]. It is lower than in our study (15.2% at 1 year), which can be explained by the EBC MAIN study requiring only a clinical follow-up, while our patients underwent angiographic follow-up at 1 year as well.
In our study, the rate of scafold thrombosis was 2.2%, and no thrombosis occurred with the metal stent. In the twostent subgroup of the EXCEL study, the rate of defnite or probable stent thrombosis at 3 years was 2.8% [14]. In the DKCRUSH-V trial, the rate of defnite or probable stent thrombosis at 3 years was 0.4% [3].
In our study, the rate of myocardial infarction was 2.2%. A higher rate of myocardial infarction (12.8%) was reported at 3 years in the two-stent subgroup of the EXCEL study [14].
Te 0% mortality rate and one instance of stent thrombosis were found encouraging. Further, predictors of TLR, the major event rate driver, were associated primarily with inadequate lesion preparation, a BVS stent size of less than 2.5 mm, and a lack of intravascular imaging. Tese results reinforce our belief that adequate lesion preparation and comprehensive imaging are essential to good long-term outcomes with BVS in the ostial LCX.

Journal of Interventional Cardiology 5
Early experience with ABSORB BVS has generally demonstrated good procedural safety and angiographic success, as well as short to mid-term clinical outcomes and safety. However, the AIDA trial showed a higher defnite and probable BVS thrombosis compared with DES, which was associated with more MI events but no signifcant diference in target vessel failure, death, or revascularization [14,15]. In addition, the recently reported ABSORB II with a 3-year follow-up found signifcantly worse outcomes regarding device-oriented composite endpoints for Absorb BVS compared with Xience DES (10% versus 5%). Interestingly, neither angina status nor coronary functional vasomotion appeared to be superior in patients with Absorb BVS compared with metallic stents. Tese fndings led to an FDA safety alert, followed by an announcement from the manufacturer regarding the discontinuation of normal sales [19]. Te major concern of BVS is late and very late scafold thrombosis [20]. However, in our study, only one early BVS   Journal of Interventional Cardiology stent thrombosis occurred (30 days after stent implantation), and no device thrombosis was observed through the 4-year follow-up. Data from the Absorb Japan trial [21] showed that, unlike standard DES, under and nonuniform device expansion was associated with greater negative remodeling and late lumen loss in ABSORB BVS, and this may in part account for the poorer outcomes after ABSORB BVS implantation compared with DES implantation in the lesions with suboptimal device expansion. Our series suggests that aggressive pre-dilatation and post-dilatation with the application of intracoronary imaging gives acceptable efcacy and safety using the BVS for complex lesions such as an unprotected left main bifurcation. Note that the one stent thrombosis occurred 40 days after the index procedure in a patient whose LCX lesion was not pretreated with a cutting balloon and the BVS diameter was 2.5 mm-both of which were predictors of the primary outcome. Te idea of "leaving nothing behind" after PCI is an intriguing concept for the treatment of bifurcation lesions.
Our study suggests there may be a place for BVS in this lesion subset and further studies should be undertaken. [22].

Limitations.
Limitations of the study include a small sample size, no control arm, and not all patients undergoing angiographic follow-up.

Conclusions
In conclusion, the use of a hybrid 2 stent/scafold strategy with DES in the main branch and BVS in the side branch in selected patients with LM true bifurcation disease was technically possible and was reasonably safe and efective. Te incidence of major events at short and long-term follow-up was similar to the published series of treatments with a two DES stent strategy. Te number of patients and events in our cohort was small with no randomized control group, so frm conclusions cannot be made. Furthermore, larger scale and randomized studies are required.

Data Availability
All data that support the fndings of this study are available from the corresponding author upon reasonable request.

Disclosure
Prof. Erglis has received institutional research grant support and advisory fees from Abbott Vascular and Boston Scientifc.

Conflicts of Interest
Te authors declare that they have no conficts of interest.