Author + information
- Received January 18, 2013
- Revision received March 22, 2013
- Accepted March 28, 2013
- Published online July 1, 2013.
- Mamoru Toyofuku, MD∗,
- Takeshi Kimura, MD†∗ (, )
- Takeshi Morimoto, MD‡,
- Yasuhiko Hayashi, MD§,
- Nobuo Shiode, MD§,
- Hideo Nishikawa, MD⋮,
- Koichi Nakao, MD¶,
- Kinya Shirota, MD#,
- Kazuya Kawai, MD∗∗,
- Yoshikazu Hiasa, MD††,
- Kazushige Kadota, MD‡‡,
- Yoichi Nozaki, MD§§,
- Takaaki Isshiki, MD⋮⋮,
- Takahito Sone, MD¶¶,
- Kazuaki Mitsudo, MD‡‡,
- j-Cypher Registry Investigators
- ∗Wakayama Medical Center, Wakayama, Japan
- †Kyoto University, Kyoto, Japan
- ‡Center for General Internal Medicine and Emergency Care, Kinki University School of Medicine, Osaka, Japan
- §Tsuchiya General Hospital, Hiroshima, Japan
- ⋮Mie Heart Center, Mie, Japan
- ¶Saiseikai Kumamoto Hospital, Kumamoto, Japan
- #Matsue Red Cross Hospital, Matsue, Japan
- ∗∗Chikamori Hospital, Kochi, Japan
- ††Tokushima Red Cross Hospital, Komatsushima, Japan
- ‡‡Kurashiki Central Hospital, Kurashiki, Japan
- §§Hokko Memorial Hospital, Sapporo, Japan
- ⋮⋮Teikyo University Hospital, Tokyo, Japan
- ¶¶Ogaki Municipal Hospital, Ogaki, Japan
- ↵∗Reprint requests and correspondence:
Dr. Takeshi Kimura, Department of Cardiovascular Medicine, Kyoto University Graduate School of Medicine, 54 Shogoin Kawahara-cho, Sakyo-ku, Kyoto 606-8507, Japan.
Objectives This study assessed 5-year outcomes after implantation of sirolimus-eluting stents (SES) for unprotected left main coronary artery (ULMCA) disease in comparison with that for non-left main disease.
Background More information on long-term outcomes after ULMCA stenting is needed.
Methods The j-Cypher is a multicenter prospective registry of consecutive patients undergoing SES implantation in Japan.
Results Among 12,812 patients enrolled in the j-Cypher registry, the unadjusted mortality rate at 5 years was significantly higher in patients with ULMCA stenting than in patients without ULMCA stenting (22.8% vs. 14.1%; p < 0.0001); however, the risk for death with ULMCA stenting was no longer significant after adjusting for confounders (hazard ratio: 1.18, 95% confidence interval: 0.95 to 1.46; p = 0.14). In the lesion-level comparison, the nonbifurcation ULMCA lesions treated exclusively with SES had a significantly lower rate of target lesion revascularization (TLR) than those in non-ULMCA nonbifurcation lesions (2.4% vs. 12.7%; p = 0.04). Among bifurcation lesions, those treated with a provisional 2-stent approach had similar rates of TLR (12.1% vs. 11.4%; p = 0.79) between the ULMCA and non-ULMCA groups. Lesions treated with an elective 2-stent approach had higher TLR rates in the ULMCA group as compared with the non-ULMCA group (33.5% vs. 19.7%; p = 0.002).
Conclusions The safety of ULMCA stenting relative to non-LMCA stenting was maintained through 5 years follow-up. In terms of efficacy, SES implantation in nonbifurcation ULMCA lesions was associated with an extremely low cumulative incidence of TLR, whereas the elective 2-stent approach for ULMCA bifurcation lesions was associated with a markedly higher cumulative incidence of TLR as compared with that for non-ULMCA bifurcation lesions.
Current guidelines allow percutaneous coronary intervention (PCI) for unprotected left main coronary disease (ULMCA) in limited situations (1,2). However, appropriate indications for PCI in patients with ULMCA are still unclear, and more information on long-term outcomes after ULMCA stenting is needed.
Our previous report from the j-Cypher registry at 3 years suggested: 1) percutaneous treatment for left main disease was safe, with a similar adjusted mortality rate as compared with the treatment for non-left main disease through 3 years; and 2) there is heterogeneity of outcomes between left main lesion locations (ostial/shaft and bifurcation) and bifurcation stenting strategies (bifurcation 1-stent and 2-stent strategies) (3).
In the current report, we assessed the longer-term safety and efficacy of ULMCA PCI relative to non-ULMCA PCI through 5 years in the j-Cypher registry cohort. The subgroup analysis comparing the outcomes of ULMCA lesions with those of non-ULMCA lesions was also conducted according to lesion locations and bifurcation stenting strategies to assess the issues specific for left main lesions.
Study design and patient population
The study design and patient enrollment for the j-Cypher registry have already been described in detail (4). In brief, the j-Cypher registry is a physician-directed, prospective, multicenter registry in Japan enrolling consecutive patients undergoing sirolimus-eluting stent (SES) implantation. From August 2004 to November 2006, 12,824 patients were enrolled in the j-Cypher registry for the first time. Excluding 2 patients without successfully treated lesions, 4 patients without stent use, and patients with bare-metal stent use only, the current study population consists of 12,812 patients treated with at least 1 SES. The recommended antiplatelet regimens were indefinite use of aspirin (81 to 100 mg/day) and a thienopyridine (200 mg of ticlopidine or 75 mg of clopidogrel daily) for at least 3 months. The duration of dual-antiplatelet therapy was left to the discretion of each attending physician. The relevant review boards in all 37 participating centers approved the study protocol. Written informed consent was obtained from all patients enrolled.
Among 12,812 study patients, 582 patients underwent PCI for ULMCA disease (ULMCA group), and 12,230 patients underwent PCI only for non-ULMCA lesions (non-ULMCA group). Lesion-based analysis was also conducted in 476 ULMCA lesions treated exclusively by SES in comparison with 14,898 non-ULMCA lesions treated similarly (Fig. 1). Lesions were divided into 3 groups including nonbifurcation lesions, bifurcation lesions treated with a provisional 2-stent approach, and bifurcation lesions treated with an elective 2-stent approach. In an attempt to select the lesions comparable with the ULMCA lesions, the following lesions were excluded from the control group: non-SES treatment, treatment with both SES and other drug-eluting stents, protected left main lesions, bifurcation lesion morphology without proximal main branch disease (Medina class 0.1.0, 0.0.1, 0.1.1), and unsuccessfully treated lesions.
The left main coronary artery was defined as “unprotected” when no surgical grafts to the left coronary system were patent. Renal failure was defined as an estimated glomerular filtration rate <30 ml/min/1.73 m2. Coronary angiographic parameters were assessed in each participating center by either visual assessment or quantitative angiographic measurement. Bifurcation lesion was defined as a lesion involving a side branch ≥2.2 mm in diameter that could be a candidate for stent implantation. When stenting was performed for the side-branch ostium (left circumflex [LCX] in the vast majority of the ULMCA cases]) before stenting of the main branch, the procedure was regarded as an elective 2-stent strategy. The provisional 2-stent strategy was defined as either main branch stenting alone or stenting for the side-branch ostium after stenting of the main branch. Bifurcation 2-stent treatment was defined as stenting of both the main and side branches, and 1-stent treatment as stenting of the main branch alone. The decision for performing a final kissing balloon dilation was also left to the operators. During the follow-up, death was regarded as cardiac in origin unless obvious noncardiac causes were identified. Any death during the index hospitalization for stent implantation was regarded as cardiac death. Myocardial infarction (MI) was adjudicated according to the definition in the Arterial Revascularization Therapy Study (5). Stent thrombosis (ST) was defined according to the Academic Research Consortium (ARC) definition (6). Target-lesion revascularization (TLR) was defined as either PCI or coronary artery bypass graft surgery due to restenosis or thrombosis of the target lesion that included the proximal and distal edge segments and the ostium of side branches. A scoring system used in the SYNTAX (Synergy Between Percutaneous Coronary Intervention With TAXUS and Cardiac Surgery) trial (SYNTAX score) was calculated to evaluate the anatomic complexity, and the tertile ranges in the original SYNTAX trial (7)were used in the post hoc analysis.
The primary outcome measure for the current analysis was all-cause death through 5 years after the index procedures. Cardiac death, MI, ST, and repeated revascularization were also assessed on a patient-level basis. TLR and ARC definite ST were evaluated on a lesion-level basis as subgroup analyses according to the lesion location and stenting strategies.
Categorical variables are presented as counts and/or percentages and were compared with the chi-square test. Continuous variables were expressed as mean ± SD or median and interquartile ranges. Continuous variables were compared with the Student ttest, analysis of variance, or Wilcoxon rank sum test on the basis of their distributions. Cumulative incidences of adverse events were estimated by the Kaplan-Meier method, and curves were compared with the log-rank test. We used the Cox proportional hazard model to make adjusted comparison for death and cardiac death. All variables in Table 1were used as potential risk-adjusting variables dichotomized by clinically meaningful thresholds and used as potential risk-adjusting variables, and we selected those variables with a univariate p value <0.05 for the multivariable model (Online Table 1, Model 1). As a sensitivity analysis to identify the fair combination of the explanatory variables, a backward variable selection on the multivariable Cox proportional hazard model with all selected factors was conducted and identified the independent explanatory variables with a p value <0.05 (Online Table 1, Model 2). Adjusted survival curves were drawn for the 2 groups of patients with or without ULMCA stenting by use of the Cox proportional hazard model in conjunction with methods described by Ghali et al. (8)with adjustment for the aforementioned variables.
On an assumption with proportional hazard by the plot of log (time) versus log [−log (survival)] stratified by independent variables, multivariable Cox proportional hazard models for TLR were developed within the first year and beyond 1 year separately, with previously identified independent risk factors consistent with our prior report (9). In evaluating risk factors for TLR beyond 1 year, we included only those patients who completed the 1-year follow-up without TLR.
Probability was considered to be significant at a level of <0.05. All statistical tests were 2-tailed. Statistical analyses were conducted by a physician (M.T.) and by an independent statistician (T.M.) with the use of JMP version 8.0.2 (SAS Institute, Cary, North Carolina), and SAS version 9.2 (SAS Institute) software.
Baseline and procedural characteristics
Patients in the ULMCA group were significantly older and had more comorbidities than those in the non-ULMCA group, as reflected by the higher incidences of stroke, heart failure, renal failure, unstable angina, shock, and bifurcation disease (Table 1).
Outcomes of patients with or without ULMCA stenting
The cumulative incidences of all-cause death and cardiac death up to 5 years were significantly higher in patients with ULMCA disease (Table 2, Fig. 2A). After adjusting for confounders, there were no significant differences in the risk for all-cause death and cardiac death between the ULMCA group and the non-ULMCA group (hazard ratio [HR]: 1.18, 95% confidence interval [CI]: 0.95 to 1.46; p = 0.14, and HR: 1.01, 95% CI: 0.72 to 1.41; p = 0.97, respectively) (Fig. 2B, Online Table 1, Model 1) As a sensitivity analysis, we conducted a backward-selection procedure on the multivariable Cox proportional hazard model with all selected risk factors, and the model yielded a similar result for all-cause death (HR: 1.18, 95% CI: 0.95 to 1.47; p = 0.13) (Online Table 1, Model 2).
Distributions of the causes of death were comparable between patients with and without ULMCA treatment (Online Table 2). There was no excess of patients who died from documented ventricular fibrillation and/or patients with sudden death in the ULMCA group.
The cumulative incidences of ST (both definite and definite/probable) were marginally higher in the ULMCA group as compared with the non-ULMCA group in the patient-level analysis (Table 2); however, the cumulative incidences of definite ST in the lesion-level analysis were not significantly different between the ULMCA and non-ULMCA groups (1.7% vs. 1.1%; p = 0.38). The cumulative incidences of TLR at the patient level were significantly higher in the ULMCA group than that in the non-ULMCA group (22.6% vs. 17.3%; p < 0.0001); however, the difference between the 2 groups was attenuated at the lesion level (15.2% vs. 13.6%; p = 0.051). The cumulative incidences of MI, stroke, and any repeated revascularization in the patient-level analysis were similar between the ULMCA and non-ULMCA groups (Table 2).
Clinical outcomes among nonbifurcation lesions
Among 476 ULMCA lesions treated exclusively with SES, 96 (20%) had ostial/shaft left main lesions and were compared with 13,249 nonbifurcation non-ULMCA lesions. With regard to lesion and procedural characteristics, ULMCA lesions had more severe calcification, more frequent IVUS guidance, shorter stent length, larger stent diameter, and higher maximal balloon inflation pressure (Table 3). Clinical outcomes at 5 years were remarkably more favorable in the ULMCA group than in the non-ULMCA group (definite ST: 0% vs. 1.2%; p = 0.35, and TLR: 2.4% vs. 12.7%; p = 0.04) (Table 4, Fig. 3). After adjusting for confounders, the lower risk for TLR in the ULMCA group was marginal (HR for early TLR within 1 year 0.27, 95% CI: 0.02 to 1.21; p = 0.1, and beyond 1 year, HR: 0.46, 95% CI: 0.07 to 1.43; p = 0.21).
Clinical outcomes among bifurcation lesions treated with the provisional 2-stent approach
Among 380 ULMCA bifurcation lesions, 281 lesions (74%) were treated with the provisional 2-stent approach, whereas 99 lesions (26%) were treated with the elective 2-stent approach. Crossover to final 2-stent treatment occurred in 20 lesions (7%) of 281 lesions with the provisional 2-stent approach. Among bifurcation lesions with the provisional 2-stent approach, the ULMCA group (281 lesions) as compared with the non-ULMCA group (1,433 lesions) had a higher prevalence of heart failure, renal failure, and more severe calcifications (Table 3). There were more frequent IVUS guidance, shorter stent length, larger stent diameter, higher balloon inflation pressure, and more frequent final kissing balloon inflation in the ULMCA lesions. The cumulative incidence of definite ST at 5 years was low in both the ULMCA and non-ULMCA bifurcation lesions treated with the provisional 2-stent approach (1.2% vs. 1.4%; p = 0.62) (Table 4, Fig. 4). The cumulative incidence of TLR was comparable between the ULMCA and non-ULMCA groups (12.1% vs. 11.4%; p = 0.79). The risk of the ULMCA lesions for TLR remained insignificant after adjusting for confounders (HR for early TLR within 1 year 1.45, 95% CI: 0.78 to 2.52; p = 0.23, and beyond 1 year, HR: 1.12, 95% CI: 0.59 to 1.98; p = 0.71).
Clinical outcomes among bifurcation lesions treated with elective 2-stent approach
Among 316 bifurcation lesions treated with the elective 2-stent approach (ULMCA 99 lesions, and non-ULMCA 217 lesions), renal failure, dialysis, and severe calcification were more prevalent in the ULMCA group than in the non-ULMCA group (Table 3). Stent diameters were larger in both main and side branches, and final kissing balloon inflation was more frequently conducted in the ULMCA group as compared with non-ULMCA group. The cumulative incidence of TLR through 5 years was significantly higher in the ULMCA group as compared with the non-ULMCA group (33.5% vs. 19.7%; p = 0.002) (Table 4, Fig. 5). After adjusting for confounders, the higher risk of ULMCA lesions for early TLR was significant (HR for TLR within 1 year 2.88, 95% CI: 1.54 to 5.44; p = 0.001, and HR for TLR beyond 1 year 0.61, 95% CI: 0.17 to 1.79; p = 0.38). Among bifurcation lesions, the interaction probability between ULMCA treatment and the elective 2-stenting for early TLR within 1 year was 0.03. The cumulative incidence of definite ST was comparable between the ULMCA and non-ULMCA groups (3.4% vs. 2.3%; p = 0.44).
Cumulative 5-year incidences of definite ST and TLR in lesions managed with bifurcation 1-stent treatment were similar between the ULMCA and non-ULMCA groups (definite ST: 1.4% vs. 1.4%; p = 0.72, and TLR: 12.3% vs. 11.0%; p = 0.67, respectively). In lesions managed with the bifurcation 2-stent treatment of both branches, the cumulative incidence of definite ST was similar between the ULMCA and non-ULMCA groups (2.8% vs. 2.1%; p = 0.56), whereas the cumulative incidence of TLR was significantly higher in the ULMCA group than in the non-ULMCA group (29.6% vs.19.7%; p = 0.007).
Risk stratification among patients treated for ULMCA
Among 476 patients treated for ULMCA exclusively with SES, the cumulative incidences of death, cardiac death, sudden death, and MI through 5 years were comparable among the 3 groups of nonbifurcation, bifurcation provisional 2-stent approach, and elective 2-stent approach (Table 5). The cumulative incidence of TLR was low in the nonbifurcation lesion, and high in bifurcation elective 2-stent treatment, as described in the previous text. The rates of TLR for each bifurcation 2-stent technique were 16 of 67 (24%) after T stenting, 9 of 32 (28%) after Culotte stenting, 4 of 15 (27%) after crush stenting, and 3 of 5 (60%) after kissing stenting. Among 302 patients in whom the SYNTAX score was available, the cumulative incidences of all-cause death, cardiac death, ST, and TLR were significantly higher in patients with high SYNTAX scores ≥33 (Online Table 3).
In this large-scale, multicenter registry in Japan, the safety of the SES treatment for ULMCA disease seems to be maintained through 5 years, with the adjusted mortality rate similar to that in patients treated for non-ULMCA disease. At lesion-level analysis, SES implantation for nonbifurcation ULMCA lesions was associated with excellent 5-year outcomes, with 2.4% of the TLR rate and no occurrence of ST. In bifurcation ULMCA lesions, the outcome of the provisional 2-stent approach was comparable to that in the non-ULMCA lesions, whereas a higher risk for TLR was observed in ULMCA lesions treated with the elective 2-stent approach as compared with non-ULMCA lesions treated similarly.
Safety of ULMCA stenting
In terms of the mortality rate after the treatment of ULMCA disease, PCI and coronary artery bypass grafting (CABG) have shown comparable outcomes in randomized control trials (10)and registries (11)for 1 to 5 years. In the current study, although no definite conclusion can be made regarding mortality at 5 years because of a relative small number of patients with a 5-year follow-up, there was no significant difference in the 5-year risk for all-cause death after adjustment for confounders between the 2 groups of patients with or without ULMCA stenting. Most of the fatal events in the ULMCA group were secondary to clinical presentation at the time of the index stent implantation and to comorbidities rather than to the performance of the implanted device. Distributions of the causes of death were comparable between patients with and without ULMCA treatment (Online Table 2). Although ST in ULMCA lesions might masquerade as sudden death, there was no excess of patients who died from documented ventricular fibrillation and/or with sudden death in the ULMCA group through 5 years. The occurrence of fatal events related to stent failure such as thrombosis or restenosis after ULMCA stenting is rare during 5 years follow-up. However, risk stratification in the patient selection for ULMCA stenting would still be required. In the subanalysis of the SYNTAX trial in the left main stratum, the mortality rate was similar between the PCI and CABG groups. In patients with SYNTAX scores ≥33, however, the cumulative incidence of all-cause death was higher in the PCI group as compared with that in the CABG group (10). In the current analysis, the cumulative incidence of 5-year mortality was markedly increased with increasing coronary anatomic complexities as assessed by the SYNTAX score (Online Table 3). Although treatment of left main disease with stenting seemed to be generally safe through the 5-year period, individualized risk stratification considering coronary anatomic complexities would be essential when choosing PCI as an alternative to CABG in the treatment of ULMCA disease.
Lesion location and bifurcation stenting strategies seemed to influence TLR outcome, but not mortality outcome (Table 5). In the nonbifurcation ULMCA disease, the risk of TLR and the cumulative incidence of ST tended to be lower compared with that in the non-ULMCA group. These favorable results were consistent with previous studies (12).
In bifurcation lesions treated with the provisional 2-stent approach, the risk for TLR and the cumulative incidence of ST were comparable between the ULMCA and non-ULMCA groups. In bifurcation lesions treated with the elective 2-stent approach, on the other hand, TLR outcome was worse in the ULMCA group than in the non-ULMCA group with a similar definite ST outcome. Although the long-term outcome of TLR for the ULMCA lesion has not yet been adequately addressed, repeated PCI for in-stent restenosis of SES was reported to be associated with a high rate of recurrence (13). Major improvements in stent design, technique, and pharmacology are needed before extending the indication for unprotected left main stenting in lesions requiring an elective 2-stent approach. Therefore, in selecting PCI for patients with ULMCA bifurcation lesions, it seems crucial to predict the ultimate need for stenting of the ostium of LCX. The very low rate of crossover from the provisional side-branch stenting approach to the final 2-stent approach in the current study was remarkable, suggesting that experienced operators could successfully choose those bifurcation lesions that did not need stenting of the ostium of the LCX. The very high rate of TLR in association with the elective 2-stent approach for the ULMCA bifurcation lesion warrants the need for further investigation.
First, the present study was an observational study. Baseline differences of patient and lesion characteristics might have biased the comparison even after adjustment for confounders. Moreover, there were no protocol-specified strategies for bifurcation stenting, and selection bias at the operator level possibly influenced the results. Secondly, because protocol-specified measurement of myocardial biomarkers was not conducted after the index procedures, we could not address the concerns for increased incidence of periprocedural non–Q-wave MI in the 2-stent group (14). Thirdly, TLR events in the present study included both clinically driven and angiographically driven events. Routine follow-up angiography was performed in most of the participating centers. The clinical significance of angiographically driven TLR of the main branch as well as the jailed side branch of ULMCA remains unclear (15). In addition, we might have been more prone to performing TLR for restenosis of ostial LCX than for restenosis of other side branches, because the former usually subtends a larger area of myocardium in jeopardy than the latter. Finally, the SYNTAX score could not be evaluated in all the patients in the ULMCA group.
The safety of ULMCA stenting relative to non-LMCA stenting seemed to be maintained through the 5-year period. In terms of efficacy, SES implantation in nonbifurcation ULMCA lesions was associated with an extremely low TLR rate, whereas the elective 2-stent approach for ULMCA lesions was associated with a markedly higher incidence of TLR as compared with that for non-ULMCA bifurcation lesions.
The authors gratefully thank the members of the cardiac catheterization laboratories of the participating centers, as well as the clinical research coordinators.
For supplementary tables, please see the online version of this paper.
This work was supported by Cordis Cardiology Japan and Johnson & Johnson. Dr. Kimura has served on the advisory board of Terumo Japan and Abbott Vascular. Dr. Isshiki is a member of the advisory boards of Abbott Vascular Japan, and Boston Scientific Japan; and has received honoraria from Abbott Vascular Japan, Boston Scientific Japan, and Terumo. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary artery bypass grafting
- confidence interval
- hazard ratio
- left circumflex coronary artery
- myocardial infarction
- percutaneous coronary intervention
- sirolimus-eluting stent(s)
- stent thrombosis
- target-lesion revascularization
- unprotected left main coronary artery
- Received January 18, 2013.
- Revision received March 22, 2013.
- Accepted March 28, 2013.
- Wijns W.,
- Kolh P.,
- Danchin N.,
- et al.
- Levine G.N.,
- Bates E.R.,
- Blankenship J.C.,
- et al.
- Toyofuku M.,
- Kimura T.,
- Morimoto T.,
- et al.
- Kimura T.,
- Morimoto T.,
- Nakagawa Y.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Kimura T.,
- Morimoto T.,
- Nakagawa Y.,
- et al.
- Morice M.C.,
- Serruys P.W.,
- Kappetein A.P.,
- et al.
- Park D.W.,
- Seung K.B.,
- Kim Y.H.,
- et al.
- Valgimigli M.,
- Malagutti P.,
- Rodriguez-Granillo G.A.,
- et al.
- Abe M.,
- Kimura T.,
- Morimoto T.,
- et al.
- Katritsis D.G.,
- Siontis G.C.,
- Ioannidis J.P.
- Koh J.S.,
- Koo B.K.,
- Kim J.H.,
- et al.