Original Contribution

Clinical and Angiographic Outcomes Comparison of Patients With Left Main Vs Non-Left Main Bifurcation Lesions Treated With Percutaneous Coronary Intervention With Second-Generation Drug-Eluting Stents

Ruben Vergara, MD;  Renato Valenti, MD;  Angela Migliorini, MD;  Michele Ciabatti, MD;  Maria Grazia De Gregorio, MD;  Giulia Taborchi, MD;  Sofia Morini, MD;  Roberto Rondine, MD;  Raffaelle Martone, MD;  David Antoniucci, MD

Ruben Vergara, MD;  Renato Valenti, MD;  Angela Migliorini, MD;  Michele Ciabatti, MD;  Maria Grazia De Gregorio, MD;  Giulia Taborchi, MD;  Sofia Morini, MD;  Roberto Rondine, MD;  Raffaelle Martone, MD;  David Antoniucci, MD

Abstract: Introduction. It is not clear if differences exist about treating left main bifurcation (LMB) and non-left main bifurcation (non-LMB) lesions by means of percutaneous coronary intervention (PCI). Methods. We prospectively analyzed all consecutive patients treated at our center for bifurcation lesions from January 1, 2011 to December 31, 2015, including acute myocardial infarction (MI) and cardiogenic shock, and compared the angiographic and clinical outcomes of patients with LMB and non-LMB lesions treated with PCI and second-generation drug-eluting stent (2G-DES) implantation. The primary endpoint was the major adverse cardiac event (MACE) composite, including MI, clinically indicated target-vessel revascularization (TVR), and cardiac death (CD) at 2-year follow-up. We also compared the angiographic patency of the vessel, which was a composite of the restenosis-reocclusion (RR) rate. Results. Out of 1081 patients (1368 bifurcations), a total of 320 patients had LMB (29%). Overall, procedural success was 98.4%. Clinical follow-up rate was 100%. Angiographic follow-up rate was 83.7%. No differences were seen regarding the primary endpoint of all MACE (17.8% in LMB vs 18.0% in non-LMB; P>.99), MI rate (4.3% in LMB vs 2.9% in non-LMB; P=.20), and CD (8.7% in LMB vs 5.8% in non-LMB; P=.08). The overall RR rate was 11.8%, with 5% RR rate in the LMB group (16/320 lesions) and 9.7% RR rate in the non-LMB group (102/1048 lesions); P<.01. The LMB group had a better TVR rate (5.0% vs 9.4% in the non-LMB group; P=.01). Conclusion. PCI with 2G-DES for LMB has better target-vessel patency and TVR rates when compared with non-LMB lesions, without clinical differences in terms of 2-year clinical outcomes.

J INVASIVE CARDIOL 2018;30(12):443-446.

Key words: coronary bifurcation lesions, drug-eluting stents, left main bifurcation, percutaneous coronary intervention

From its introduction in 2008, second-generation drug-eluting stent (2G-DES) implantation has demonstrated an improvement in the rates of myocardial infarction (MI) and stent thrombosis (ST) when compared with first-generation (1G)-DES and has become the first choice for coronary stenting.1-3 In the same sense, the treatment of unprotected left main (LM) lesions with percutaneous coronary intervention (PCI) with 2G-DES has consistently improved long-term clinical outcomes.4 However, PCI for distal LM lesions remains challenging. The approach and result accepted for LM bifurcations (LMBs) should be different when compared with non-LMBs. Although it may be satisfactory to accept a suboptimal result with Thrombolysis in Myocardial Infarction (TIMI) flow grade 1 in a small side branch (SB), such a result is not acceptable when treating a distal LMB.5 However, few data exist comparing LMB vs non-LMB patients treated with PCI in the 2G-DES era. In this study, we sought to determine if differences exist regarding clinical and angiographic outcomes of real-world patients treated with PCI and 2G-DES for LMB vs non-LMB lesions.


We performed a prospective analysis of all consecutive patients treated for bifurcation lesions at our center from January 1, 2011 to December 31, 2015. Exclusion criteria were SB minimal diameter <2 mm or LM disease already treated with coronary artery bypass graft surgery. Patients presenting with acute coronary syndrome (ACS), acute MI, and cardiogenic shock were included. The primary endpoint was the major adverse cardiac event (MACE) composite, including clinically indicated target-vessel revascularization (TVR), MI, and cardiac death (CD) at 2-year follow-up. We also compared the target-vessel patency rate, which was defined as a composite of the restenosis-reocclusion (RR) rate.

PCIs were performed according to current standard guidelines. Patients with LMB were treated with PCI due to presenting ACS or after evaluation by the hospital’s heart team. All patients underwent oral dual-antiplatelet therapy for at least 12 months. In cases of acute MI, we systematically used an intravenous bolus of abciximab unless contraindicated. Patients then received a loading dose of prasugrel 60 mg or ticagrelor 180 mg orally and continued treatment for at least 12 months. Intravenous heparin was administered to maintain an activated clotting time of 250-300 seconds. The vascular access route, type of 2G-DES (everolimus, zotarolimus, biolimus, or amphilimus), and decision to treat bifurcation lesions with 1-stent or 2-stent technique were left to the operator’s discretion.

All patients were scheduled for 6-9 month angiographic follow-up if not contraindicated by renal insufficiency in asymptomatic patients, and/or clinical follow-up at 6 months and yearly thereafter. TVR was performed on the basis of significant restenosis or reocclusion with symptoms, inducible ischemia, and/or segmental ventricular dysfunction. In order to better evaluate the potential differences between LMB and non-LMB patients, we also performed a comparison dividing the groups by lesion treated.

Statistical analysis. Discrete data were summarized as frequencies, whereas continuous data were summarized as mean ± standard deviation or median and interquartile range (IQR) when appropriate. The Fisher’s exact test or Pearson’s Chi-square test were used for comparison of categorical variables, and unpaired and paired 2-tailed Student’s t-tests were used to detect differences among continuous variables. Survival curves were generated using the Kaplan-Meier method, and the difference between groups was assessed by a log-rank test. A multivariable Cox regression forward stepwise analysis was performed to analyze independent variables correlated with the primary endpoint. The variables included in the model were age, diabetes mellitus, multivessel disease, previous MI, left ventricular ejection fraction <30%, LMB treatment, multiple bifurcation disease, ACS at presentation, and restenosis of the vessel treated at baseline PCI. All tests were two-sided, and a P-value <.05 was considered statistically significant. The analysis was performed with SPSS software, version 19 (IBM).


Out of 1081 patients treated for bifurcations (overall 1368 lesions), a total of 320 (29%) had LMB and 438 patients (40%) were treated because of ACS. A total of 255 patients (23.5%) were treated for >1 bifurcation. Three patients (1 in the LMB group and 2 in the non-LMB group) died during the index procedure (P>.99). Table 1 shows the baseline clinical characteristics. The group of patients with LMB had a significant major risk profile: they were older and more often diabetic, with higher rates of renal insufficiency and severe ventricular dysfunction. The LMB group had a significantly higher rate of multiple bifurcations involved and treated. On the other hand, ACS was more often the indication for the index PCI in the non-LMB group. There was no difference regarding the overall PCI success rate between groups. 

Table 2 shows the procedural characteristics. Of note, the Medina classification 1,1,1 was prevalent in the LMB group, while 1,1,0 was more frequent in the non-LMB group. Lesion length was significantly longer in the non-LMB group. The use of intravascular ultrasound guidance was significantly higher in the LMB group (72.0% vs 28.9% in the non-LMB group; P<.001), while there was a higher incidence of stent implantation in both the main and side branch in the LMB group (23.7% vs 18.8% in the non-LMB group; P=.08). There was a high use of kissing-balloon technique in both groups. Fluoroscopy time and contrast volume were slightly higher in the LMB group. Overall procedural success was 98.6%, with no difference between the two groups. 

There were 15 deaths during the index recovery (4.6%) in the LMB group, and 25 deaths (3.3%) in the non-LMB group (P=.22). Among the 1041 patients who survived, the 2-year clinical follow-up rate was 100%. Two patients (1 in each group) had definite ST, and 5 patients (2 in the LMB group and 3 patients in the non-LMB group) had probable ST (P>.99). The angiographic follow-up rate was 80.4% (778/971 eligible patients). Tables 3 and 4 illustrate the angiographic follow-up results stratified by patients and lesions treated, respectively. The overall RR rate was 11.8% (118/997 lesions with angiographic follow-up).

There was no difference between the groups for the primary composite endpoint of all MACE: 17.8% (57 patients) in the LMB group vs 18.0% (137 patients) in the non-LMB group (P>.99). The reinfarction rate was 4.3% (14 patients) in the LMB group vs 2.9% (22 patients) in the non-LMB group (P=.20). The cardiac death rate was 8.7% in the LMB group (28 patients, of which 18 [64.0%] had multiple bifurcation lesions) while 5.8% (44 patients) died in the non-LMB group (P=.08). 

The independent predictors correlated with the primary endpoint at the Cox Regression analysis were: age >75 years (hazard ratio [HR], 1.671; 95% confidence interval [CI], 1.252-2.229; P<.001), diabetes mellitus (HR, 1.543; 95% CI, 1.141-2.086; P<.01), and left ventricular ejection fraction <30% (HR, 1.727; 95% CI, 1.203-2.481; P<.01).


The results of this study can be summarized as follows: (1) PCI with 2G-DES implantation for coronary bifurcations has a high rate of procedural success and optimal clinical and angiographic follow-up, beyond the anatomic segment involved; (2) Only clinical variables are independent predictors of poor cardiovascular outcome; and (3) LMB lesions have better RR rates than non-LMB lesions.

Poor data exist from randomized clinical trials (RCTs) and observational studies comparing long-term outcomes in patients treated for bifurcation lesions with PCI and 2G-DES. Most RCTs have been performed only in non-LMB and focused on 1-stent vs 2-stent strategies.6,7 Data are also very poor regarding the comparison between LMB vs non-LMB lesions. One registry with a large number of patients in this setting was a sub-analysis of COBIS II (Coronary Bifurcation Stenting Registry in South Korea) that compared techniques for LMB vs non-LMB lesions, in which >80% of patients were treated with 1G-DES.8 The authors suggested that LMB could have worse clinical outcomes. In our study, using only 2G-DES implantation, there was a high procedural success rate regardless of the anatomic segment treated, with no differences in clinical outcomes between the LMB and non-LMB groups at 2 years. When comparing the groups by lesions treated, the RR and TVR rates were higher in the non-LMB group. These rates are similar to those described in some RCTs and other registries in this setting6-8 and are not correlated with poor clinical outcomes.

While most studies focused on 1G-DES,9 data are very poor regarding the rate of late ST in bifurcation lesions treated with 2G-DES. In our study, there was a very low rate of definite or probable ST, with no difference between the groups. We hypothesize several procedural reasons for this: all patients underwent long-term dual-antiplatelet therapy; the high rate of preparation of the lesion before stenting (thrombectomy, predilation, rotational atherectomy); use of 2G-DES; and the high rate of non-compliant kissing-balloon postdilation in both groups. 

In our analysis, the LMB group had a better rate of long-term target-vessel patency, but there were several significant procedural differences between the groups; in particular, the shorter lesion length, more frequent use of IVUS guidance, and fewer thrombus-containing lesions in the LMB group all favor the angiographic outcomes in this group.

Finally, the treatment of bifurcation lesions remains technically challenging. Bifurcations represent up to 20% of coronary stenoses and are associated with increased long-term risk of reintervention and ST compared with non-branching lesions.10 A relatively recent analysis of the Korean Bifurcation Pooled Cohorts found that the independent predictors of 3-year cumulative incidence of target-vessel failure for coronary bifurcation lesions with 2-stent strategy were LMB, high SYNTAX (Synergy Between Percutaneous Coronary Intervention With Taxus stents and Cardiac Surgery) score, and diabetes mellitus, while 2G-DES, use of non-compliant balloon, and final kissing balloon were independent variables of better outcomes.11 Although the studies are not comparable, our results are similar except for the LMB treatment. Age >75 years, diabetes mellitus, and left ventricular ejection fraction <30% were independent variables of worse outcomes.

Study limitations. This was a non-randomized study; however, the multivariable Cox regression analysis confirmed that LMB treatment was not an independent variable correlated with worse prognosis. Treatment strategy and stent type were left to the operator’s discretion, which may influence differences in angiographic outcomes between the groups. Despite that limitation, this is a real-world study showing the differences in the contemporary technical approach to both LMB or non-LMB lesions. 


PCI with 2G-DES implantation for bifurcation lesions has high procedural success rate, with good clinical and angiographic outcomes. PCI with 2G-DES for LMB lesions has better target-vessel patency and TVR rates when compared to non-LMB lesions, without clinical differences in terms of 2-year clinical outcomes.


1.    Navarese EP, Kowalewski M, Kandzari D, et al. First-generation versus second-generation drug-eluting stents in current clinical practice: updated evidence from a comprehensive meta-analysis of randomized clinical trials comprising 31 379 patients. Open Heart. 2014;1:e000064.

2.    Collet JP, Silvain J, Kerneis M, et al; for the ARCTIC Investigators. Clinical outcome of first- vs second-generation DES according to DAPT duration: results of ARCTIC-generation. Clin Cardiol. 2016;39:192-200.

3.    Stefanini GG, Baber U, Windecker S, et al. Safety and efficacy of drug-eluting stents in women: a patient-level pooled analysis of randomized trials. Lancet. 2013;382:1879-1888. Errata in Lancet. 2013;382:1878. 

4.    Stone GW, Sabik JF, Serruys PW, et al; for the EXCEL Trial Investigators. Everolimus-eluting stents or bypass surgery for left main coronary artery disease. N Engl J Med. 2016;375:2223-2235.

5.    Latib A, Colombo A. Bifurcation disease: what do we know, what should we do? JACC Cardiovasc Interv. 2008;1:218-226.

6.    Grundeken MJ, Wykrzykowska JJ, Ishibashi Y, et al. First generation versus second generation drug-eluting stents for the treatment of bifurcations: 5-year follow-up of the LEADERS all-comers randomized trial. Catheter Cardiovasc Interv. 2016;87:E248-E260.

7.    Niemelä M, Kervinen K, Erglis A, et al. Nordic-Baltic PCI Study Group. Randomized comparison of final kissing balloon dilatation versus no final kissing balloon dilatation in patients with coronary bifurcation lesions treated with main vessel stenting: the Nordic-Baltic Bifurcation Study III. Circulation. 2011;123:79-86.

8.    Song YB, Hahn JY, Yang JH, et al. Differential prognostic impact of treatment strategy among patients with left main versus non-left main bifurcation lesions undergoing percutaneous coronary intervention: results from the COBIS (Coronary Bifurcation Stenting) Registry II. JACC Cardiovasc Interv. 2014;7:255-263.

9.    Zimarino M, Corazzini A, Ricci F, Di Nicola M, De Caterina R. Late thrombosis after double versus single drug-eluting stent in the treatment of coronary bifurcations. A meta-analysis of randomized and observational studies. JACC Cardiovasc Interv. 2013;6:687-695.

10.    Chen SL, Louvard Y, Runlin G. Perspective on bifurcation PCI. J Interv Cardiol. 2009;22:99-109.

11.    Song PS, Song YB, Lee JM, et al. Major predictors of long-term clinical outcomes after percutaneous coronary intervention for coronary bifurcation lesions with 2-stent strategy: patient-level analysis of the Korean bifurcation pooled cohorts. JACC Cardiovasc Interv. 2016;9:1879-1886.

From the Division of Cardiology, Careggi Hospital, Florence, Italy.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted July 5, 2018, provisional acceptance given July 20, 2018, final version accepted July 25, 2018.

Address for correspondence: David Antoniucci, MD, Division of Cardiology, Careggi Hospital, Viale Pieraccini 17, I-50134, Florence, Italy. Email: ruben.arg@gmail.com