Long-Term Outcomes of Different Two-Stent Techniques With Second-Generation Drug-Eluting Stents for Unprotected Left Main Bifurcation Disease: Insights From the FAILS-2 Study

Abstract: Objectives. To investigate the long-term clinical outcomes of second-generation drug-eluting stent (2G-DES) implantation for the treatment of complex unprotected left main coronary artery (ULMCA) bifurcation lesions with different two-stent techniques. Background. Several two-stent techniques for ULMCA bifurcation lesions have been described. However, a paucity of data exists regarding the optimal strategy, especially in the 2G-DES era. Methods. The FAILS-2 registry enrolled 1270 consecutive patients treated for ULMCA stenosis with 2G-DES. We compared long-term outcomes of different two-stent strategies in patients who underwent PCI for complex ULMCA bifurcation disease. The primary endpoints were the incidence of death and major adverse cardiac events (MACE, defined as a composite of all-cause death, myocardial infarction [MI], target-lesion revascularization [TLR], and stent thrombosis [ST]) at long-term follow-up. Results. A total of 238 patients were included in the present analysis. T-stenting strategy was used in 66 patients, mini-crush in 104 patients, and culotte in 68 patients. After a median follow-up of 2.27 years, death rates were comparable for the three techniques (9.3% T-stenting vs 9.0% mini-crush vs 4.5% culotte [P=.48]). MACE rates were also similar between the three groups (22% T-stenting vs 26% mini-crush vs 31% culotte [P=.50]). Finally, we showed no differences in MI, ST, and TLR rates between groups. At multivariate analysis, no significant advantage of one technique over the others was observed. Conclusion. T-stenting, mini-crush, and culotte techniques using 2G-DES for ULMCA bifurcation disease showed similar clinical outcomes at long-term follow-up. MACE rates were mainly driven by in-stent restenosis at the circumflex ostium.

J INVASIVE CARDIOL 2018;30(8):276-281

Key words: culotte technique, two-stent strategies, LM bifurcation disease, mini-crush technique, T-stenting technique

Significant unprotected left main coronary artery (ULMCA) disease is reported in nearly 4%-6% of patients undergoing coronary angiography,¹ and most of these lesions (65%-80% of cases) are bifurcations.^2,3 Drug-eluting stent (DES) implantation has significantly improved the outcomes of PCI patients in comparison with bare-metal stents.^4,5 Second-generation (2G)-DES options with or without reabsorbable polymers provide better outcomes in comparison to early-generation DESs.^6-9

Although a number of studies on LM bifurcation lesions have demonstrated that treatment with a single stent is superior to a two-stent strategy,^10-12 30%-70% of patients^13,14 required a planned two-stent strategy. Actually, only a few studies evaluated the impact of different stenting strategies to treat LM bifurcation lesions using first-generation (1G)-DES options, and to date there are no randomized trials from the 2G-DES era. FAILS-2 is a multicenter registry of patients treated with 2G-DES for ULMCA lesions.³ The present substudy investigates the long-term outcomes of different two-stent PCI strategies with 2G-DES for patients with ULMCA bifurcation lesions. 

Methods

Study population. The methods of the FAILS-2 registry were published previously.³ In brief, the registry includes all-comer patients with ULMCA disease treated with 2G-DES between January 2007 and January 2015 at six centers in Europe and Japan (Città della Scienza e della Salute di Turin, Italy; Ospedale di Rivoli, Turin, Italy; Ospedale San Giovanni Bosco, Turin, Italy; San Raffaele Scientific Institute and EMO-GVM Columbus, Milan, Italy; New Tokyo Hospital, Matsudo, Japan; and Hospital Clinico San Carlos, Madrid, Spain).     

Of the 1270 patients who underwent PCI with 2G-DES, 304 patients (24%) were treated with two-stent technique for distal LM bifurcation disease. Stenting strategy was not adequately defined or attributable in 66 patients, who were thus excluded from the present analysis. A total of 238 patients represented our study population: 217 (91%) presented with a true bifurcation lesion, defined by a Medina classification of (1,1,1), (1,1,0), (1,0,1), or (0,1,1), while 21 patients (9%) presented with a non-true bifurcation lesion, defined by a Medina classification of (1,0,0), (0,1,0), or (0,0,1). In all cases, the selected revascularization approach seemed suitable to guarantee complete revascularization. All data related to hospital admissions, procedures, and outcomes were collected at each center’s hospital recording network. Information regarding the clinical status at the latest clinical follow-up available was collected by clinical visits, telephone interviews, and referring physician interviews. Routine programmed angiographic follow-up was left to the discretion of the enrolling center, while angiographic follow-up was performed if a non-invasive evaluation or clinical presentation suggested ischemia. Data analysis was performed with the approval of the institutional ethics committees of the hospitals involved.

Definitions. LM bifurcation lesion was defined as a lesion at the distal segment of the LM with Medina classification of (1,1,1), (1,1,0), (1,0,0), (1,0,1), (0,1,1), (0,0,1), or (0,1,0). In our analysis, we collected only patients with ULMCA bifurcation lesions who underwent 2G-DES implantation, like previously published in the FAILS-2 registry. ULMCA was defined as LM stenosis and no patent bypass graft to the left coronary system (left anterior descending [LAD] or left circumflex (LCX) arteries).  

Antiplatelet therapy and periprocedural anticoagulation followed standard regimens. All procedures were performed with standard interventional techniques; the use of predilation, intra-aortic balloon pump, type of two-stent technique and 2G-DES types were selected at the treating physician’s discretion.    

Preprocedural and postprocedural intravascular ultrasound (IVUS) was used according to each center’s protocol to obtain more accurate information on the disease status of the distal LM complex for the selection of the most appropriate and safest stenting technique.

Patients were divided by two-stent strategy into three groups: (1) T-stenting group; (2) mini-crush group; and (3) culotte group. Postdilation with a non-compliant balloon was performed in cases of suboptimal stent expansion at angiographic or IVUS imaging or routinely at the center’s discretion. 

Renal insufficiency was defined as estimated glomerular filtration rate <30 mL/min/1.73 m² according to the Modification of Diet in Renal Disease study equation. Deaths were classified as either cardiac or non-cardiac. Cardiac death was defined as any death due to a cardiac cause (eg, myocardial infarction [MI], low-output failure, fatal arrhythmia) and procedure-related deaths. Target-lesion revascularization (TLR) was defined as any repeat intervention of the target segment or in the adjacent proximal or distal 5 mm. Target-vessel revascularization (TVR) was defined as any repeat intervention of any segment of the target vessel, defined as the entire major coronary vessel proximal and distal to the target lesion, including upstream and downstream branches and the target lesion. Spontaneous MI was defined as the occurrence after hospital discharge of any value of troponin and/or CK-myocardial band greater than the upper limit of normal if associated with clinical and/or electrocardiogram change. Ischemia-driven TLR was defined as any PCI or coronary artery bypass graft (CABG) surgery occurring at the target segment (or in the adjacent proximal or distal 5 mm) for angina or angina-equivalent symptoms, with or without documented ischemia. Ischemia-driven TVR was defined as any PCI or CABG surgery occurring at any segment of the target vessel for angina or angina-equivalent symptoms, with or without documented ischemia. Stent thrombosis (ST) was defined on the basis of the Academic Research Consortium definitions. Major adverse cardiac event (MACE) was defined as the composite endpoint of death, MI, and TVR. 

Statistical analysis. Categorical variables are reported as counts and percentages, whereas continuous variables as mean ± standard deviation or interquartile range (IQR). Gaussian or non-Gaussian distribution was evaluated by Kolmogorov-Smirnov test. The Student’s t-test was used to assess differences between parametric continuous variables, the Mann-Whitney U-test was used for non-parametric variables, the Chi-square test was used for categorical variables, and Fisher’s exact test was used for 2 x 2 tables. Cox multivariate analysis was performed to assess the independent predictors of MACE. A two-sided P-value <.05 was considered statistically significant; all analyses were performed with SPSS 21.0 (IBM, Inc).

Study endpoints. The primary endpoints were the incidence of death and composite MACE at long-term follow-up.

Results

Baseline characteristics (Table 1). The study patients were grouped by the different two-stent techniques: T-stenting (n = 66), mini-crush (n = 104), and culotte (n = 68). There were no differences in cardiovascular risk factor rates between the different groups.  Mean EuroSCORE values were 4.3 ± 1.4 for T-stenting vs 4.6 ± 2.4 for mini-crush vs 4.7 ± 2.4 for culotte patients (P=.81). SYNTAX scores ≥33 were found in 42% of the T-stenting group, 37% of the mini-crush group, and 31% of the culotte group (P=.46).

Procedural characteristics (Table 2). Femoral access was the preferred revascularization site in all three groups; coronary angiography showed a high prevalence of true bifurcation disease (86.4% in T-stenting group, 93.3% in mini-crush group, 92.6% in culotte group; P=.44), in particular of Medina bifurcation 1,1,1 (56% of the T-stenting group, 64.5% of the mini-crush group, and 67.5% of the culotte group; P=.35). IVUS was performed in 17% of the T-stenting group, 37.5% of the mini-crush group, and 32% of the culotte group (P=.01). Optical coherence tomography and fractional flow reserve were rarely performed.

Everolimus-eluting stent was the most widely used stent (84% T-stenting vs 60% mini-crush vs 66% culotte).

At quantitative coronary analysis (QCA), stent diameter was smaller with T-stenting and mini-crush technique vs the Culotte technique (3.1 ± 0.7 mm vs 3 ± 0.7 mm vs 3.6 ± 0.5 mm, respectively; P<.01), and stent length was shorter (18 ± 6 mm vs 19 ± 6 mm vs 21 ± 5 mm, respectively; P<.01). High-pressure postdilations were performed less frequently for T-stenting group for either main branch (77.2% vs 92.3% vs 94.1%; P<.01) and side branch (59% vs 82% vs 88%; P<.01) for the T-stenting, mini-crush, and culotte groups, respectively. Final kissing balloon was achieved in 85% vs 92% vs 94% of the T-stenting, mini-crush, and culotte groups, respectively (P=.07). 

In-hospital and long-term outcomes (Table 3). During hospitalization, only 1 death was registered, in the mini-crush group (P=.52). Planned angiographic follow-up was performed in 50%, 68%, and 79.5% of the T-stenting, mini-crush, and culotte groups, respectively (P<.01). Long-term follow-up (median follow-up for the overall population, 2.7 years) was achieved for the majority of patients. All-cause death rate (9.3% vs 9.0% vs 4.5%; P=.48) were not significantly different between groups. No ST occurred, and MI rates were 1.5% vs 1.0% vs 1.4% for the T-stenting, mini-crush, and culotte groups, respectively (P=.93).

Rates of TVR (P=.28) and TLR (P=.41) were not different between groups, even if they were numerically higher both in the mini-crush group (19.5% and 17.5%, respectively) and culotte group (25.0% and 20.5% respectively) than in the T-stenting group (14.0% and 12.5%, respectively), which was probably secondary to higher planned angiographic follow-up rates. Moreover, the site of in-stent restenosis was more frequently the CX ostium (9.3% vs 15.5% vs 16% for the T-stenting, mini-crush, and culotte groups, respectively; P=.44).

Thus, MACE rates at follow-up were not different among groups (22% vs 26% vs 31% for the T-stenting, mini-crush, and culotte groups, respectively; P=.50). Kaplan-Meier curves for MACE, all-cause death, TLR, and CX ostium TLR at follow-up are illustrated in Figure 1.

In the multivariable Cox regression analysis adjusted for covariates, no significant advantage of one technique over the others could be observed, while male gender (hazard ratio [HR], 0.506; 95% confidence interval [CI], 0.27-0.93; P=.03) and stable angina at presentation (HR, 0.52; 95% CI, 0.29-0.92; P=.02) resulted protective (Table 4). 

Discussion

This study on 2G-DES implantation in patients treated with different two-stent strategies for ULMCA bifurcation lesions demonstrated: (1) extremely low rates of in-hospital mortality; (2) no differences in long-term outcomes using different two-stent strategies; (3) MACE mainly driven by TLR at the CX artery ostium; and (4) no definite or probable ST observed. 

Even if a number of studies and randomized trials on bifurcation coronary lesions have demonstrated that single-stent strategy is superior to two-stent strategies, including with the use of newer-generation DESs,^5,15-17 the recently published DK CRUSH-V trial showed a lower 1-year target-lesion failure (TLF) rate for PCI of true LM bifurcation lesions (1,1,1; 0,1,1) using a planned DK CRUSH two-stent strategy vs a provisional strategy.¹⁸ However, in this trial, 47.1% of patients in the provisional stenting group underwent side-branch bailout stenting for suboptimal results after main-vessel stenting; in this setting, the delivery of the side-branch stent could have been complicated by imprecise placement, incomplete expansion or asymmetry, or edge dissection that could have contributed to higher TLF rate at medium-term follow-up in the provisional stenting group. 

Previous data on different two-stent strategies in the 1G-DES era showed conflicting results. In the DK-CRUSH III study,¹⁹ the culotte technique vs double-kissing crush was associated with significantly increased MACE rates (23.7% vs 8.2%, respectively; P<.001) at 3-year follow-up. On the contrary, an observational registry¹⁵ showed similar clinical outcomes in patients with different treatment approaches, with MACE-free survival rates of 66.5% in the T-stenting group, 69.3% in the V-stenting group, and 66.9% in the crush-stenting group at 2-year follow-up. Another 1G-DES study²⁰ demonstrated that even if MACE rates were non-statistically different between mini-crush and culotte (36% vs 41%), MI (0% vs 11.3%; P<.01) and ST rates (0% vs 6.3%; P=.02) were significantly higher with culotte technique at 5-year follow-up. 

Among patients who underwent 2G-DES implantation in the ISAR-LEFT MAIN-2 trial,²¹ a two-stent strategy was planned for 233 patients; 80% of patients were treated with culotte, 17% with T-stenting, and 3% with crush techniques. The authors stated that no interaction between stenting technique and treatment effect of the two-stent types was observed. 

In the present study on patients treated with 2G-DES, different two-stent technique were associated with an extremely low in-hospital mortality (0% vs 0.9% vs 0% for T-stenting, mini-crush, and culotte, respectively) in spite of severe lesion complexity. In addition, the long-term outcomes with 2G-DES for ULMCA disease seemed to be acceptable. The rate of MI was extremely low (0.4% of the overall population, which is lower than previous reports from the 1G-DES era).^15,19,20 In contrast with the 1G-DES era, where complex stenting for bifurcation disease was independently related to ST,¹² no definite or probable ST was reported at long-term follow-up in the present study.^19,20 

MACE rates were comparable among the T-stenting, mini-crush, and culotte groups at long-term follow-up (22% vs 26% vs 31%, respectively; P=.50); as with previous IG-DES studies, these rates were mainly driven by TLR on the CX artery ostium.^6,14,20,22 This may be explained by the high rates of planned angiographic follow-up, which might have over-estimated the severity of the LCX ostium, and resulted in the high rate of CX-TLR. The liberal use of physiological assessments for LCX ostium at follow-up could be useful in order to reduce unnecessary revascularizations, resulting in better clinical outcomes. 

This multicenter study also reflects the current practice when treating ULMCA bifurcation lesions. In the main paper of the FAILS-2 registry, we demonstrated that IVUS-guided LM-PCI was protective in terms of TLR on the LM at follow-up and in previous reports seemed to be associated with reduction in mortality²³ and TLR²⁴ even though this has not been evaluated in randomized trials. The MAIN-COMPARE (Revascularization for Unprotected Left Main Coronary Artery Stenosis: Comparison of Percutaneous Coronary Angioplasty vs Surgical Revascularization) registry revealed a significant reduction of the cumulative mortality in the IVUS-guided PCI group compared to the angiography-guided PCI group (4.7% vs 16%; P=.048).²³ It might be reasonable that IVUS use has more impact to reduce MACE following planned two-stent strategy compared to single crossover stenting.

It is important to note that a two-stent technique for ULMCA lesions should be distinguished from other bifurcation lesions because of their clinical importance, mechanical and anatomical features (ie, hinge motion of the LCX ostium). Furthermore, once the side branch (LCX) is acutely occluded, this can be a fatal complication. In addition, a specific two-stent technique cannot be applied for all true bifurcation lesions. The operators must consider the vessel size of both branches, bifurcation angle, and severity of the side-branch stenosis. Unfortunately, we could not investigate further details of anatomical characteristics in the distal LM lesions (ie, bifurcation angle, vessel diameters of the main and side branches, etc). The selection of two-stent technique was dependent on operator discretion at each institute. Therefore, our results cannot be generalized for every LM lesion. Randomized trials are eagerly awaited based on our results.

Conclusion

Despite greater lesion complexity in this substudy, 2G-DES two-stent strategies for ULMCA disease suggests acceptable in-hospital and long-term outcomes; MACE rates were mainly driven by TLR at the side branch (LCX), and no ST was registered at follow-up. Further randomized studies dedicated to the treatment of true LM bifurcation are required.

References

1.    Ragosta M. Left main coronary artery disease: importance, diagnosis, assessment, and management. Curr Probl Cardiol. 2015;40:93-126. Epub 2014 Nov 14. 

2.    Lassen JF, Holm NR, Banning A, et al. Percutaneous coronary intervention for coronary bifurcation disease: 11th consensus document from the European Bifurcation Club. EuroIntervention. 2016;12:38-46. 

3.    D’Ascenzo F, Chieffo A, Cerrato E, et al. Incidence and management of restenosis after treatment of unprotected left main disease with second-generation drug-eluting stents (from Failure in Left Main Study With 2nd Generation Stents – Cardiogroup III Study). Am J Cardiol. 2017;119:978-982. Epub 2017 Jan 6.

4.    Park SJ, Kim YH, Lee BK, et al. Sirolimus-eluting stent implantation for unprotected left main coronary artery stenosis: comparison with bare metal stent implantation. J Am Coll Cardiol. 2005;45:351-356.

5.    Kim YH, Park DW, Lee SW, et al. Long-term safety and effectiveness of unprotected left main coronary stenting with drug-eluting stents compared with bare-metal stents. Circulation. 2009;120:400-407. 

6.    Moynagh A, Salvatella N, Harb T, et al. Two-year outcomes of everolimus vs. paclitaxel-eluting stent for the treatment of unprotected left main lesions: a propensity score matching comparison of patients included in the French Left Main Taxus (FLM Taxus) and the LEft MAin Xience (LEMAX) registries. EuroIntervention. 2013;9:452-462. 

7.    Buchanan GL, Chieffo A, Bernelli C, et al. Two-year outcomes following unprotected left main stenting with first vs new-generation drug-eluting stents: the FINE registry. EuroIntervention. 2013;9:809-816. 

8.    Valenti R, Migliorini A, Parodi G, et al. Clinical and angiographic outcomes of patients treated with everolimus-eluting stents or first-generation paclitaxel-eluting stents for unprotected left main disease. J Am Coll Cardiol. 2012;60:1217-1222. 

9.    Park SJ, Ahn JM, Kim YH, et al. Temporal trends in revascularization strategy and outcomes in left main coronary artery stenosis: data from the ASAN Medical Center-Left MAIN Revascularization registry. Circ Cardiovasc Interv. 2015;8:e001846. 

10.    Hildick-Smith D, de Belder AJ, Cooter N, et al. Randomized trial of simple versus complex drug-eluting stenting for bifurcation lesions: the British Bifurcation Coronary Study: old, new, and evolving strategies. Circulation. 2010;121:1235-1243.

11.    Behan MW, Holm NR, Curzen NP, et al. Simple or complex stenting for bifurcation coronary lesions: a patient-level pooled-analysis of the Nordic Bifurcation Study and the British Bifurcation Coronary Study. Circ Cardiovasc Interv. 2011;4:57-64. Epub 2011 Jan 4.

12.    Colombo A, Bramucci E, Saccà S, et al. Randomized study of the crush technique versus provisional side-branch stenting in true coronary bifurcations: the CACTUS (Coronary Bifurcations: Application of the Crushing Technique Using Sirolimus-Eluting Stents) study. Circulation. 2009;119:71-78. Epub 2008 Dec 22.

13.    Kim YH, Park DW, Ahn JM, et al. Everolimus-eluting stent implantation for unprotected left main coronary artery stenosis. The PRECOMBAT-2 (Premier of Randomized Comparison of Bypass Surgery versus Angioplasty Using Sirolimus-Eluting Stent in Patients with Left Main Coronary Artery Disease) study. JACC Cardiovasc Interv. 2012;5:708-717. 

14.    Tiroch K, Mehilli J, Byrne RA, et al. Impact of coronary anatomy and stenting technique on long-term outcome after drug-eluting stent implantation for unprotected left main coronary artery disease. JACC Cardiovasc Interv. 2014;7:29-36.

15.    Palmerini T, Marzocchi A, Tamburino C, et al. Impact of bifurcation technique on 2-year clinical outcomes in 773 patients with distal unprotected left main coronary artery stenosis treated with drug-eluting stents. Circ Cardiovasc Interv. 2008;1:185-192. 

16.    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. 

17.    Toyofuku M, Kimura T, Morimoto T, et al. Three-year outcomes after sirolimus-eluting stent implantation for unprotected left main coronary artery disease: insights from the j-Cypher registry. Circulation. 2009;120:1866-1874.

18.    Chen SL, Zhang JJ, Han Y, et al. Double kissing crush versus provisional stenting for left main distal bifurcation lesions: DKCRUSH-V randomized trial. J Am Coll Cardiol. 2017;70:2605-2617.

19.    Chen SL, Xu B, Han YL, et al. Comparison of double kissing crush versus culotte stenting for unprotected distal left main bifurcation lesions: results from a multicenter, randomized, prospective DKCRUSH-III study. J Am Coll Cardiol. 2013;61:1482-1488.

20.    Kawamoto H, Takagi K, Chieffo A, et al. Long-term outcomes following mini-crush versus culotte stenting for the treatment of unprotected left main disease: insights from the Milan and New-Tokyo (MITO) registry. Catheter Cardiovasc Interv. 2017;89:13-24. 

21.    Mehilli J, Richardt G, Valgimigli M, et al. Zotarolimus- versus everolimus-eluting stents for unprotected left main coronary artery disease. J Am Coll Cardiol. 2013;62:2075-2082. 

22.    Takagi K, Ielasi A, Shannon J, et al. Clinical and procedural predictors of suboptimal outcome after the treatment of drug-eluting stent restenosis in the unprotected distal left main stem: the Milan and New-Tokyo (MITO) registry. Circ Cardiovasc Interv. 2012;5:491-498. 

23.    Park SJ, Kim YH, Park DW, et al; MAIN-COMPARE Investigators. Impact of intravascular ultrasound guidance on long-term mortality in stenting for unprotected left main coronary artery stenosis. Circ Cardiovasc Interv. 2009;2:167-177. 

24.    Kang SJ, Ahn JM, Song H, et al. Comprehensive intravascular ultrasound assessment of stent area and its impact on restenosis and adverse cardiac events in 403 patients with unprotected left main disease. Circ Cardiovasc Interv. 2011;4:562-569.

From the ¹Città della Salute e della Scienza di Torino, Italy; ²Ospedale degli Infermi di Rivoli, Turin, Italy; ³New Tokio Hospital, Matsudo, Japan; ⁴San Raffaele Scientific Institute, Milan, Italy; ⁵Hospital Clinico San Carlos, Madrid, Spain; and ⁶Ospedale San Giovanni Bosco, Turin.

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 March 12, 2018 and accepted March 26, 2018.

Address for correspondence: Marco Pavani, MD, Division of Cardiology, Città della Salute e della Scienza di Torino, Italy. Email: marcopavani@alice.it. Website: www.cardiogroup.org