Unprotected Left Main Coronary Artery Disease: Outcomes of Treatment With Second-Generation Drug-Eluting Stents – Insight From the FAILS-2 Study

Abstract: Objective. To evaluate the outcome of patients undergoing PCI for unprotected left main coronary artery (ULMCA) disease with different drug-eluting stent (DES) types. Background. Published literature suggests that second-generation DES options have differing vascular responses and outcomes, but there is a paucity of data in real-life patients in the LM setting. Methods. This is a retrospective, multicenter study, including patients treated with a second-generation DES for ULMCA disease between 2007 and 2015. The primary endpoint was target-lesion revascularization (TLR). Secondary endpoints were major adverse cardiac events, myocardial infarction (MI), and stent thrombosis (ST). Results. A total of 1209 patients were enrolled; 840 patients (69.5%) received an everolimus-eluting stent (EES), 133 patients (11.0%) received a zotarolimus-eluting stent (ZES), and 236 patients (19.5%) received a biodegradable polymer, biolimus-eluting stent (BP-BES). During a mean follow-up of 722 ± 640 days, TLR occurred in 47 patients (3.8%). At univariate analysis, EES patients had a lower TLR rate (3.6% vs 4.5% in ZES vs 4.2% in BP-BES), which was statistically significant at multivariate analysis (hazard ratio, 0.50; 95% confidence interval, 0.27-0.93; P=.03). No differences in major adverse cardiac events, death, MI, or ST were observed between groups. Conclusion. The safety profile of the stents used was comparable over the follow-up period. However, EES patients had lower restenosis rates, with a reduced need for repeat PCI. 

J INVASIVE CARDIOL 2018;30(8):283-288. Epub 2018 May 15.

Key words: left main coronary artery disease, comparative effectiveness/patient-centered outcomes research, drug-eluting stent

Unprotected left main coronary artery (ULMCA) disease carries a poor prognosis in those untreated.^1-5 For many years, surgical revascularization has been the mainstay of treatment, favored over PCI. This primarily was due to the potential risk of stent thrombosis (ST) and restenosis – the latter of significant concern in the era of bare-metal stents, with reports suggesting restenosis rates of 15%-60% due to intrinsically peculiar anatomic features of the ULMCA.⁶ However, developments in stents and eluting polymer technologies have been incorporated in new-generation drug-eluting stent (DES) options that optimize the process of vessel healing and thus avoid these complications,^7-9 leading to new guideline indications.¹⁰

In the early 2000s, the SYNTAX trial was the first to randomize to percutaneous coronary intervention (PCI) or coronary artery bypass graft (CABG) surgery in an all-comers design for the treatment of multivessel coronary disease.¹¹ The subgroup analysis highlighted no difference in major adverse cardiac event (MACE) rate at 5 years between those treated with either CABG or PCI with Taxus stents (Boston Scientific) for ULMCA disease.¹² While SYNTAX was limited by the use of Taxus stents, the newer-generation DES options allow for more efficient vessel healing post stent implantation without compromising luminal diameter.⁷ Their biocompatible or resorbable polymers reduce inflammatory responses and late ST.^13,14

However, few data on real-life patients have been available. NOBLE and EXCEL, two randomized trials comparing CABG vs PCI for ULMCA using second-generation DESs showed conflicting results,^15,16 which were also due to differences in follow-up duration and stent types used. Both old and recent data suggest that these stents have differing vascular responses and outcomes, and it would not be inconceivable to postulate the superiority of one stent over the other in this setting of LM disease.⁷ Our group, therefore, aimed to compare the effectiveness of different second-generation DES options in this setting. 

Methods

This retrospective, multicenter study recruited patients undergoing PCI to critical ULMCA between June 2007 and January 2015. Relevant patients were identified from a multicenter registry,¹⁷ and dedicated follow-up was performed in all patients to note well-being, as well as primary and secondary endpoints (discussed later). This was either direct (through physical or telephone assessment) or indirect (through inquiries with respective primary-care physicians).

Once relevant patients had been identified from the registry, clinical data, including follow-up, were collected from electronic patient records. Angiographic data from the initial index PCI and any subsequent angiographic information were collected after at least two dedicated physicians had reviewed all imaging determining the lesion site(s), classifications, and calculation of SYNTAX scores.¹¹ For this analysis, we excluded all patients with no available data regarding stent type/dimensions (Figure 1).

Coronary angioplasty and stent implantation during index PCI was performed according to established practices and guidelines at the time of each procedure. The choices regarding device, technique (including the approach to bifurcation stenting, kissing balloons, and postdilation), and drug therapy (including glycoprotein IIb/IIIa inhibitors) were at the operator’s discretion. Post procedure, all patients were prescribed aspirin (100 mg lifelong) and a P2Y12 inhibitor. The choice between additional angiographic or sole clinical follow-up was at the discretion of the operator and referring physician, taking into account patient preference and comorbidities. In the majority of cases, angiographic and clinical follow-up was recommended 6-12 months after the index PCI, irrespective of symptoms or signs of ischemia. 

Treatment of restenosis was also at the operator’s discretion, but collectively discussed in the majority of cases, and final management was based on the patient’s symptoms and/or signs of ischemia (including fractional flow reserve and instantaneous wave-free ratio evaluation at the moment of angiography), coronary anatomy, surgical risk, feasibility of PCI, and overall life expectancy. In cases of repeat PCI, the techniques and devices were at the operator’s discretion.

The primary endpoint for this study was target-lesion revascularization (TLR). Secondary endpoints were all-cause death, myocardial infarction (MI), target-vessel revascularization (TVR), and ST. Independent predictors of TLR were the cosecondary endpoints. 

The study size of approximately 1000 patients was decided upon after reviewing the work of Mehilli et al¹⁸ and Peduzzi et al.¹⁹ In these studies, a TLR rate of approximately 10% was quoted for Taxus stents in the LM, and at least 1000 patients were needed to test in a multivariate model. 

Statistical analysis. Categorical variables are reported as counts and percentages, while continuous variables are reported as mean and standard deviation or interquartile range (IQR). Gaussian or non-Gaussian distribution was evaluated by Kolmogorov-Smirnoff test. Student’s t-test or 3-way ANOVA were used to assess differences between parametric continuous variables, while Mann-Whitney U-test was used for non-parametric variables, 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 TLR, defined as every variables that resulted in a difference with P<.10 at univariate analysis. Three different analyses were performed to test the different effects of the DES. A two-sided P-value <.05 was considered statistically significant. All analyses were performed with SPSS 21.0 (IBM) and GraphPad Prism 6.1 (GraphPad Software, Inc).

Results

The FAILS-2 registry population included 1270 patients; we analyzed 1209 patients who received treatment for ULMCA disease with everolimus-eluting stent (EES), zotarolimus-eluting stent (ZES), or biodegradable polymer, biolimus-eluting stent (BP-BES) with complete stent data (Figure 1). Average patient age was 70.3 ± 10.3 years and 924 (76.5%) were male. Risk factors including hypertension and diabetes were found in 80.8% and 39.2%, respectively, and 41.9% had previously undergone PCI. SYNTAX score was calculated for all patients, while SYNTAX II score was available in only 737 patients (Table 1). 

The presenting complaint leading to initial angiography in the majority of patients was stable angina (49.8%). The presentation was acute in the remaining patients (37.2% with non-ST elevation acute coronary syndrome [NSTEACS], 8.8% with ST-elevation myocardial infarction [STEMI], and 7.2% with cardiogenic shock).

Xience V EES (Abbott Vascular) or Promus Premier EES (Boston Scientific) were implanted in 69.5%, Endeavor ZES or Resolute Integrity ZES (both Medtronic) were implanted in 11%, and Biomatrix Flex BP-BES (Biosensors International) or Nobori BP-BES (Terumo) were implanted in 19.5% (Figure 2). The EES group had more stable angina patients (52.1% vs 43.2% for ZES and 45.3% for BP-BES). As expected, on the other side of the coin, BP-BES had the majority of STEMI patients (20.5% vs 10.9% for ZES and 6.90% for EES). 

During a mean follow-up of 722 ± 640 days, MACE occurred in 282 patients (23.3%) and death occurred in 140 patients (11.6%). No differences were found between stent types at overall population analysis. Furthermore, no differences existed in terms of death, MI, or ST (Table 2).

However, repeat LM revascularization was lower in patients treated with EES (3.6% vs 4.5% for ZES and 4.2% for BP-BES) (Figure 3). At multivariate analysis, the use of EES was the only independent predictor of lower TLR at follow-up (hazard ratio, 0.50; 95% confidence interval, 0.27-0.93; P=.03) (Table 3, Figure 4), which may explain the above lower number of events.

Major bleeding was noted to be higher in the EES group (10.4% vs 6.7% for ZES and 2.6% for BP-BES), but the EES patients also had the lowest rate of dual-antiplatelet therapy (DAPT) interruption (16.4% vs 27.1% for ZES and 18.2% for BP-BES; P=.01) (Table 2). Furthermore, early DAPT interruption doesn’t seem to influence the LM-TLR rate (Figure 5).

Due to frequent overlapping of LM segments, no significant difference emerged about the choice of stent depending on the LM segment, although BP-BES options were less frequently used for mid-LM stenosis (Table 4). The preferred stenting technique was provisional (77.9% overall) (Table 4). The majority of patients received DAPT with acetylsalicylic acid plus clopidogrel (Table 4).

Discussion

This study was designed to assess the effectiveness of different second-generation DES options for the management of ULMCA disease. The findings of this study are particularly important because, together with existing data, they may guide stent selection in the growing area of LM-PCI.

Nevertheless, before discussing the differences in outcomes between the three stent types, it is perhaps important to discuss the overall MACE and death rates observed in our study. These will infer the efficacy of PCI as a treatment strategy in this setting. In our study, we report MACE and death rates of 23.3% and 11.6%, respectively. These are somewhat higher than previous reports. Specifically, a meta-analysis of 1278 patients published in 2008 quoted a MACE rate of 10.6%.²⁰ More recently, the SYNTAX and EXCEL studies reported similar MACE rates of approximately 15%.^11,16

However, our results are more closely matched with the NEST registry, where MACE occurred in 18.8% during a median follow-up of 551 days.²¹ In addition, our results are more favorable than ISAR-LEFT-MAIN and ISAR-LEFT-MAIN 2, which involved 1257 patients with a follow-up of 3 years and a MACE rate of 28%.^18,22

This disparity in trial results is quite significant, and may certainly be explained by a number of reasons, but one that cannot be ignored is patient selection; a close analysis of the baseline characteristics of our cohort highlights a number of unwell patients. Our patients with cardiogenic shock and STEMI, who made up 7.2% and 8.8% of our cohort, respectively, would have met the exclusion criteria of contemporary trials. Furthermore, our study population had a relatively high proportion of NSTEMIs compared with EXCEL (16.8% vs 13%, respectively).¹⁶ The recognition of higher in-hospital mortality in patients undergoing ULMCA-PCI in the setting of ACS (compared with stable angina) may partially explain our results.²³

In addition, we noted a greater proportion of complex coronary anatomy. While we do not detail location of disease in the LM and technique used to treat the lesion, we have quoted SYNTAX scores, and again noted a difference compared with other contemporary trials. Patients with high SYNTAX scores were excluded from EXCEL,¹⁶ and both NOBLE and EXCEL had mean SYNTAX scores of 22 and 20 (low), respectively – notably lower than our cohort, with a mean of 26 (moderate).^15,16 In fact, it would appear that our study is almost an outlier, as demonstrated by a recently published meta-analysis of eight trials comparing LM-PCI vs LM-CABG in 4700 patients;²⁴ in the meta-analysis, 22% of patients had SYNTAX scores of >32, while 27% had scores >32 in our study. This same analysis also highlighted a better outcome with CABG in those with higher SYNTAX scores, affirming the findings from the SYNTAX trial.¹¹ This difference in coronary complexity may well be what is giving rise to the ranges in MACE quoted in the literature. It would also explain the very non-favorable MACE rate noted in ISAR-LEFT-MAIN 2, where 47% of patients had the highest SYNTAX scores.¹⁸ Twenty-seven percent of patients included in our study underwent PCI despite a high SYNTAX score; this was due to patient choice or high surgical risk. This mirrors the real world, in which PCI patients are often in the worst general conditions (as is reflected by the higher prevalence of patients with elevated SYNTAX II scores in our groups), and it could be an explanation for the significant amount of MACE at follow-up, when only half of deaths were cardiovascular, leading to a non-cardiac death rate of >6%. 

Superiority with EES was noted for the primary endpoint of TLR. This evidently is very interesting, and pertinent in this setting. The same criticism applies as above, with fewer acute patients in this group, but published data suggest similar outcomes at 2 years after an initially higher in-hospital mortality rate for those undergoing LM-PCI in an acute setting rather than elective.^23,26 Thus, EES may well confer an advantage in terms of TLR.  

These findings are consistent with EXCEL, a large, randomized control trial,¹⁶ and may explain why NOBLE did not reproduce the findings of EXCEL, as the main stent used in that trial was the Biomatrix BP-BES.¹⁵ 

To the best of our knowledge, this is the first time EES implantation has shown superiority over other second-generation DES options in the setting of LM disease. ISAR-LEFT-MAIN 2 compared ZES and EES, but was not able to show a statistically significant result, despite a trend toward less TLR with EES.¹⁸ It may well be that the follow-up of only 12 months was not long enough. Future randomized trials are clearly required to help build a body of evidence to support the findings from our study. We may even find that rather than the drug, the platform or polymer might confer the advantage, which needs further study. To that end, an ongoing trial comparing Synergy and Xience stents for the treatment of LM disease (IDEAL-LM trial) is long awaited.²⁷

Study limitations. This study is observational and non-randomized, leaving it liable to confounding factors. As the choice of stent was based on the physicians’ discretion, the results may be affected by selection biases, but the grouping by drug family aimed to reduce these biases and to reflect the real-world stenting strategy in ULMCA disease. Additionally, the groups were not equal in size, and not entirely matched. A propensity-score match would have probably increased statistical significance, but would have dramatically reduced the number of patients included. In addition, the limited use of imaging (30% of the entire population) could have affected outcomes; nevertheless, this reflects a real-life setting. Data on ST are probably under-reported in our study, due to the ominous consequence of ST on ULMCA, which may have reduced the number of such diagnoses among patients with cardiovascular death.

Conclusion

In our registry, we demonstrated that no discernable differences were noted between stents, but EES did show superiority in terms of TLR. This is in keeping with the EXCEL trial results, and we hope our findings spark interest in further randomized trials to aid with stent selection for the treatment of ULMCA disease, while awaiting the results of IDEAL-LM.  

References

1.    D’Ascenzo F, Presutti DG, Picardi E, et al. Prevalence and non-invasive predictors of left main or three-vessel coronary disease: evidence from a collaborative international meta-analysis including 22,740 patients. Heart. 2012;98:914-919. 

2.    D’Ascenzo F, Barbero U, Cerrato E, et al. Accuracy of intravascular ultrasound and optical coherence tomography in identifying functionally significant coronary stenosis according to vessel diameter: a meta-analysis of 2,581 patients and 2,807 lesions. Am Heart J. 2015;169:663-673. 

3.    Naganuma T, Chieffo A, Meliga E, et al. Long-term clinical outcomes after percutaneous coronary intervention versus coronary artery bypass grafting for ostial/midshaft lesions in unprotected left main coronary artery from the DELTA registry: a multicenter registry evaluating percutaneous coronary intervention versus coronary artery bypass grafting for left main treatment. JACC Cardiovasc Interv. 2014;7:354-361. 

4.    D’Ascenzo F, Barbero U, Moretti C, et al. Percutaneous coronary intervention versus coronary artery bypass graft for stable angina: meta-regression of randomized trials. Contemp Clin Trials. 2014;38:51-58.

5.    Morice MC, Serruys PW, Kappetein AP, et al. Five-year outcomes in patients with left main disease treated with either percutaneous coronary intervention or coronary artery bypass grafting in the synergy between percutaneous coronary intervention with taxus and cardiac surgery trial. Circulation. 2014;129:2388-2394.

6.    Puranik AS, Dawson ER, Peppas NA. Recent advances in drug eluting stents. Int J Pharm. 2013;441:665-679.

7.    Fujino Y, Attizzani GF, Tahara S, et al. Difference in vascular response between sirolimus-eluting and everolimus-eluting stents in ostial left circumflex artery after unprotected left main as observed by optical coherence tomography. Int J Cardiol. 2017;230:284-292.

8.    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. 

9.    D’Ascenzo F, Taha S, Moretti C, et al. Meta-analysis of randomized controlled trials and adjusted observational results of use of clopidogrel, aspirin, and oral anticoagulants in patients undergoing percutaneous coronary intervention. Am J Cardiol. 2015;115:1185-1193.

10.    Windecker S, Kolh P, Alfonso F, et al. 2014 ESC/EACTS guidelines on myocardial revascularization: the Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) developed with the special contribution of the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur Heart J. 2014;35:2541-2619.

11.    Serruys PW, Morice MC, Kappetein AP, et al; for the SYNTAX Investigators. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med. 2009;360:961-972.

12.    Head SJ, Davierwala PM, Serruys PW, et al. Coronary artery bypass grafting vs percutaneous coronary intervention for patients with three-vessel disease: final five-year follow-up of the SYNTAX trial. Eur Heart J. 2014;35:2821-2830.

13.    Maupas E, Lipiecki J, Levy R, et al. Safety and efficacy outcomes of 3rd generation  DES in an all-comer population of patients undergoing PCI: 12-month and 24-month  results of the e-Biomatrix French registry. Catheter Cardiovasc Interv. 2017 May  22 (Epub ahead of print).

14.    Garot P, Morice MC, Tresukosol D, et al; LEADERS FREE Investigators. 2-year outcomes of high bleeding risk patients after polymer-free drug-coated stents. J Am Coll Cardiol. 2017;69:162-171.

15.    Mäkikallio T, Holm NR, Lindsay M, et al; NOBLE Study Investigators. Percutaneous coronary angioplasty versus coronary artery bypass grafting in treatment of unprotected left main stenosis (NOBLE): a prospective, randomized, open-label, non-inferiority trial. Lancet. 2016;388:2743-2752.

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

17.    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.

18.    Mehilli J, Richardt G, Valgimigli M, et al; ISAR-LEFT-MAIN 2 Study Investigators. Zotarolimus- versus everolimus-eluting stents for unprotected left main coronary artery disease. J Am Coll Cardiol. 2013;62:2075-2082.

19.    Peduzzi P, Concato J, Kemper E, Holford TR, Feinstein AR. A simulation study of the number of events per variable in logistic regression analysis. J Clin Epidemiol. 1996;49:1373-1379.

20.    Biondi-Zoccai GG, Lotrionte M, Moretti C, et al. A collaborative systematic review and meta-analysis on 1278 patients undergoing percutaneous drug-eluting stenting for unprotected left main coronary artery disease. Am Heart J. 2008;155:274-283.

21.    Bernelli C, Chieffo A, Buchanan GL, et al. New-generation drug-eluting stent experience in the percutaneous treatment of unprotected left main coronary artery disease: the NEST registry. J Invasive Cardiol. 2013;25:269-275.

22.    Xhepa E, Tada T, Kufner S, et al. Long-term prognostic value of risk scores after drug-eluting stent implantation for unprotected left main coronary artery: a pooled analysis of the  ISAR-LEFT-MAIN and ISAR-LEFT-MAIN 2 randomized clinical trials. Catheter Cardiovasc Interv. 2017;89:1-10.

23.    Hyun DY, Jeong MH, Sim DS, et al. Two-year clinical outcomes in stable angina and acute coronary syndrome after percutaneous coronary intervention of left main coronary artery disease. Korean J Intern Med. 2016;31:1084-1092.

24.    Mahmoud AN, Elgendy IY, Mentias A, et al. Percutaneous coronary intervention or coronary artery bypass grafting for unprotected left main coronary artery disease. Catheter Cardiovasc Interv. 2017 Mar 15 (Epub ahead of print).

25.    Jensen LO, Kaltoft A, Thayssen P, et al. Outcome in high risk patients with unprotected left main coronary artery stenosis treated with percutaneous coronary intervention. Catheter Cardiovasc Interv. 2010;75:101-108.

26.    Lemmert ME, Oldroyd K, Barragan P, et al. Reduced duration of dual antiplatelet therapy using an improved drug-eluting stent for percutaneous coronary intervention of the left main artery in a real-world, all-comer population: rationale and study design of the prospective randomized multicenter IDEAL-LM trial. Am Heart J. 2017;187:104-111.

From the ¹“SS Annunziata” Hospital, Savigliano, Italy; ²Royal Brompton & Harefield Hospital, London, United Kingdom; ³Interventional Cardiology, Infermi Hospital, Rivoli and San Luigi Gonzaga, Orbassano (Turin), Italy; ⁴“Città della Scienza e della Salute,” Department of Cardiology, University of Turin, Turin, Italy; ⁵Interventional Cardiology Unit New Tokyo Hospital, Matsudo, Japan; ⁶Department of Medico-Surgical Sciences and Biotechnologies, Sapienza University of Rome, Latina, Italy; Department of AngioCardioNeurology, IRCCS Neuromed, Pozzilli, Italy; ⁷University Hospital Zurich, Department of Cardiology, Zurich, Switzerland; ⁸“S.G. Bosco Hospital,” Department of Cardiology, Turin, Italy; ⁹Scientific Institute S. Raffaele, Department of Cardiology, Milan, Italy; and ¹⁰Hopital Clinico San Carlos, Department of Cardiology, Madrid, Spain.

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 11, 2018, final version accepted March 26, 2018.

Address for correspondence: Umberto Barbero, Invasive Cardiology Unit, “SS. Maria Annunziata” Hospital, Via Ospedali 14, Savigliano, Italy. Email: ubarbero@unito.it