Author + information
- Received June 21, 2016
- Revision received July 11, 2016
- Accepted July 14, 2016
- Published online November 14, 2016.
- Dong-Ho Shin, MD, MPH, DrPHa,
- Sung-Jin Hong, MDb,
- Gary S. Mintz, MDc,
- Jung-Sun Kim, MDa,d,
- Byeong-Keuk Kim, MDa,d,
- Young-Guk Ko, MDa,d,
- Donghoon Choi, MDa,d,
- Yangsoo Jang, MDa,d,e and
- Myeong-Ki Hong, MDa,d,e,∗ ()
- aDepartment of Internal Medicine, Severance Cardiovascular Hospital, Yonsei University Health System, Seoul, South Korea
- bDepartment of Internal Medicine, Sanggye Paik Hospital, Inje University, Seoul, South Korea
- cCardiovascular Research Foundation, New York, New York
- dCardiovascular Research Institute, Yonsei University College of Medicine, Seoul, South Korea
- eSeverance Biomedical Science Institute, Yonsei University College of Medicine, Seoul, South Korea
- ↵∗Reprint requests and correspondence:
Dr. Myeong-Ki Hong, Division of Cardiology, Severance Cardiovascular Hospital, Yonsei University College of Medicine, Yonsei-ro 50-1, Seodaemun-gu 03722, Seoul, South Korea.
Objectives The aim of this study was to evaluate the clinical usefulness of intravascular ultrasound (IVUS)–guided new-generation drug-eluting stent (DES) implantation using a meta-analysis of individual patient–level data from randomized trials.
Background Published randomized trials that compare IVUS-guided versus angiography-guided new-generation DES implantation are scarce.
Methods Searches of the MEDLINE, Embase, and Cochrane databases were performed to find randomized trials that compared IVUS-guided versus angiography-guided new-generation DES implantation. A total of 2,345 patients from 3 randomized trials were identified, and all patients were treated for long lesions or chronic total occlusions. Individual patient–level data were obtained. The primary endpoint was a major adverse cardiac event, a composite of cardiac death, myocardial infarction, or stent thrombosis. An intention-to-treat analysis and per protocol analysis were performed.
Results By 1 year post-procedure, major adverse cardiac events had occurred in 0.4% of the patients who underwent IVUS-guided DES implantation versus 1.2% of those who underwent angiography-guided DES implantation (hazard ratio [HR]: 0.36; 95% confidence interval [CI]: 0.13 to 0.99; p = 0.040). For the IVUS-guided group, favorable clinical outcomes were observed for myocardial infarction (0% vs. 0.4%; HR: 0.09; p = 0.026). In addition, the clinical benefit of IVUS guidance was stronger in the per protocol analysis (HR: 0.32; 95% CI: 0.12 to 0.89; p = 0.021).
Conclusions Compared with angiographic guidance, IVUS-guided new-generation DES implantation was associated with favorable outcomes in terms of major adverse cardiac events, the composite of cardiac death, myocardial infarction, or stent thrombosis. These findings must be interpreted only for complex lesions, because all identified patients had long lesions or chronic total occlusions.
The results of 6 recently published meta-analyses indicated that intravascular ultrasound (IVUS)–guided drug-eluting stent (DES) implantation was associated with a significant reduction in major adverse cardiac events (MACE), stent thrombosis, and target lesion revascularization (TLR) (1–6), but these studies included patients who received first-generation DES and used only study-level (not patient-level) data. In addition, 5 of these 6 meta-analyses included observational studies (1–5); the sixth study was the meta-analysis that included only randomized trial data (6).
Because clinical outcomes between first- and new-generation DES-treated patients are clearly different (7) and the first-generation DES are not currently used in daily clinical practice, data from studies with exclusively next-generation DES–treated patients are required. Furthermore, previous randomized studies failed to prove improvement of hard clinical endpoints in IVUS-guided new-generation DES implantation (8,9). According to the IVUS-XPL (Impact of Intravascular Ultrasound Guidance on Outcomes of XIENCE PRIME Stents in Long Lesions) trial, IVUS-guided everolimus-eluting stent implantation for long coronary lesions had an approximately 50% reduction in the 1-year rate of MACEs, the composite of cardiac death, myocardial infarction, or TLR (8). However, the reduction in MACEs was driven mainly by the reduction in TLR, without between-group differences in cardiac death or myocardial infarction, which may be more clinically important events. Therefore, our objective was to conduct a meta-analysis of individual patient–level randomized trial data to evaluate whether IVUS guidance improves hard clinical endpoints in new-generation DES–treated patients.
Study design and selection
The meta-analysis was performed according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement guidelines (10). The MEDLINE, Embase, and Cochrane databases were searched for randomized trials published from 2005 through 2015 that compared IVUS-guided versus angiography-guided new-generation DES implantation. The search terms were “intravascular ultrasound” and “drug-eluting stent” (Online Appendix). Only full-paper published studies were considered for inclusion in the meta-analysis; there were no language restrictions. Studies that included the use of first-generation DES or that consisted of abstract-only data or both were excluded from our study. Information on study design, inclusion and exclusion criteria, patient and procedural characteristics, and clinical outcomes was extracted. The Cochrane bias assessment tool was used by 2 of the investigators (D.-H.S. and S.-J.H.), who independently assessed the risk for bias (11). Conflicts between the 2 investigators were resolved by consensus. The final selection for inclusion in the meta-analysis consisted of 3 randomized controlled trials (8,9,12). To obtain the individual patient–level data, the study statisticians from each trial extracted the patient-level data by direct access to the study databases. The independent statistician (D.-H.S.) collected all data from the individual trials and cross-checked them against previous publications. Data on baseline patient characteristics and procedure information, and data on clinical events, were collected. These patient data were pooled and analyzed in a single dataset.
Endpoints and definitions
The primary endpoint in this meta-analysis was MACE of hard clinical endpoint, defined as a composite of cardiac death, myocardial infarction, and stent thrombosis. The secondary endpoint was individual components of the primary endpoint and TLR. Academic Research Consortium criteria were used to define clinical events (13); the specific endpoint definitions, as applied in each trial, were also incorporated into the study. All deaths were considered cardiac deaths unless a definite noncardiac cause could be established (13,14). Myocardial infarction during 1-year follow-up after hospital discharge was defined as the presence of clinical symptoms, electrocardiographic changes, or abnormal imaging findings that indicated myocardial infarction, combined with an increase in the creatine kinase myocardial band fraction above the upper normal limit or an increase in troponin T or troponin I to greater than the 99th percentile of the upper normal limit (8,9,12,13). Definite, probable, and possible stent thrombosis were defined according to the recommendations of the Academic Research Consortium (13,14). TLR was defined as repeat percutaneous coronary intervention or bypass surgery of target lesions with either of the following (according to each study): 1) ischemic symptoms or positive stress test results and angiographic diameter stenosis ≥50% measured by quantitative coronary angiographic analysis; or 2) angiographic diameter stenosis ≥70% measured by quantitative coronary angiographic analysis without ischemic symptoms or positive stress test results (8,9,12).
For the individual patient–level analysis, results for continuous variables are presented as mean ± SD. Results for categorical variables are presented as number (percentage). Continuous and categorical variable data were analyzed using Student t tests and chi-square tests, respectively. Cumulative incidence values were calculated using the Kaplan-Meier method and were compared using the log-rank test. Hazard ratios (HRs) were calculated using the Cox proportional hazards model, stratified by trial. Specifically, information from trials in which patients had enrolled was included as a random effect using a gamma frailty fit (15). Firth’s penalized method was used if there was a convergence problem caused by lack of events (16). The analysis was performed using intention-to-treat and per protocol analysis. Subgroup analysis was performed according to the baseline characteristics. Two-sided p values were used, and a p value <0.05 was considered to indicate a statistically significant result. All analyses were performed using R version 3.2.2 (The R Foundation for Statistical Computing, Vienna, Austria).
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses study flow is illustrated in Figure 1. Three randomized trials, RESET IVUS (Real Safety and Efficacy of 3-Month Dual Antiplatelet Therapy Following Endeavor Zotarolimus-Eluting Stent Implantation), CTO-IVUS (Chronic Total Occlusion Intervention With Drug-Eluting Stents Guided by Intravascular Ultrasound), and IVUS-XPL, were included in the meta-analysis (8,9,12). The study design, characteristics, and results of the 3 randomized trials are summarized in Table 1.
A total of 2,345 randomized patients were identified. These patients were randomized to undergo either IVUS-guided or angiography-guided next-generation DES implantation to treat complex coronary lesions (e.g., long lesions and chronic total occlusions). The results of the risk for bias assessment are presented in Online Figure 1. Blinding of the operator could not be performed for any of the studies, because the information from IVUS itself was the result of random allocation. Otherwise, no apparent risk for bias was found in terms of random sequence generation, allocation concealment, blinding of outcome assessment, or reporting.
The results for baseline patient characteristics and procedure information are presented in Table 2. The random allocations between the IVUS-guided and angiography-guided arms were well balanced.
The results for clinical outcomes are presented in Figure 2 and Table 3. By 1 year post-implantation, the primary endpoint (MACEs) had occurred in 0.4% of the patients who underwent IVUS-guided DES implantation versus 1.2% of those who underwent angiography-guided DES implantation (HR: 0.36; 95% confidence interval [CI]: 0.13 to 0.99; p = 0.040) (Figure 2A). In addition, the benefit of IVUS guidance was stronger in the per protocol analysis. MACEs occurred in 0.4% of the patients who underwent IVUS-guided DES implantation versus 1.3% of those who underwent angiography-guided DES implantation (HR: 0.32; 95% CI: 0.12 to 0.89; p = 0.021) (Figure 2B). Favorable clinical outcomes were observed for myocardial infarction in the patients who underwent IVUS-guided versus angiography-guided procedures (0% vs. 0.4%, respectively, p = 0.026) (Figure 2D). The occurrence of cardiac death and stent thrombosis was not significantly different between the 2 groups (Figures 2C and 2E).
The results for the subgroup analysis indicated that no statistically significant interactions were present (Figure 3). There were also no signs of between-trial heterogeneity for any of the results (Online Figure 2). A funnel plot analysis did reveal, however, that publication bias might have contributed to the result for the clinical endpoint (Online Figure 3).
This meta-analysis of 3 randomized trials included individual patient–level data from a total of 2,345 new-generation DES–treated patients. The analysis revealed that compared with angiography-guided DES implantation, IVUS-guided DES implantation for the treatment of complex coronary lesions was associated with a reduction in MACEs. Use of IVUS resulted in a relative risk reduction of 64% for MACE, the composite of cardiac death, myocardial infarction, and stent thrombosis at 1 year. These findings were consistent across the clinical (lesion and patient) subgroups and were stronger according to the per protocol analysis.
The IVUS-XPL trial was the largest randomized controlled trial performed to date. This trial revealed that 1,400 patients who underwent IVUS-guided everolimus-eluting stent implantation to treat long coronary lesions had an approximately 50% reduction in the 1-year rate of MACEs (5.8% with angiographic guidance vs. 2.9% with IVUS guidance; HR: 0.48; p = 0.007), which was due primarily to the lower risk for TLR (8). However, in that trial, the between-group differences in cardiac death or myocardial infarction were not statistically different because of the relatively low event rate in the patients treated with new-generation DES. The occurrence of cardiac death was 0.4% for the IVUS-guidance arm and 0.7% for the angiography-guidance arm (HR: 0.60; p = 0.48). Myocardial infarction occurred in only 1 patient (0.1%); this patient received an angiography-guided stent (p = 0.32) (8). In the present study, we found that the MACEs (occurrence of cardiac death, myocardial infarction, or stent thrombosis) were significantly less frequent in patients who underwent IVUS-guided DES implantation compared with those who underwent angiography-guided DES implantation (0.4% vs. 1.2%, respectively; HR: 0.36; 95% CI: 0.13 to 0.99; p = 0.040). Notably, in the present study, the primary endpoint did not include TLR. Therefore, different from the IVUS-XPL trial showing the benefit of IVUS due primarily to the less frequent TLR events, MACEs, even omitting the TLR events in this meta-analysis, were less frequent with IVUS guidance. In addition, consistent with the IVUS-XPL trial, the reduction in TLR with IVUS guidance versus angiographic guidance was also confirmed in the present meta-analysis. TLR was further reduced with IVUS guidance in the as-treated analysis. Therefore, reduction of TLR in complex lesions despite the use of new-generation DES is a major advantage of IVUS guidance.
The results of previous meta-analyses suggest that there was a clinical benefit to the use of IVUS during DES implantation (i.e., a significant reduction in MACEs, stent thrombosis, or TLR) (1–6). However, the proportions of first-generation (Cypher and Taxus) DES used in these studies were substantial. In 1 meta-analysis that included the largest number of DES-treated patients (n = 26,503), the proportion of new-generation DES–treated patients was <45% (4). In addition, most of the studies included in these meta-analyses were observational, rather than randomized controlled, studies, and significant heterogeneity was present in the analyses (1–4). The most recent updated meta-analysis did include only randomized controlled trials with 3,192 patients, but this study also included 792 patients (25%) who received first-generation DES (6). Furthermore, only a study-level analysis was performed by the investigators (6). Our study is the first meta-analysis to use individual patient–level data. It included data from 2,345 patients from randomized trials of the new-generation DES. The results indicated that the use of IVUS had clinical benefits for the hard clinical endpoint, the composite of cardiac death, myocardial infarction, and stent thrombosis.
Because the use of new-generation DES produces better clinical outcomes compared with the use of even first-generation DES, the clinical benefit of IVUS has been estimated to be smaller than previously thought (17–19). Nevertheless, a clinical benefit for the hard clinical endpoint variables was revealed by the present study. This can be partly explained in that the 3 randomized trials included in our present study enrolled patients with complex coronary lesions (e.g., chronic total occlusions and diffuse long lesions). Despite the use of new-generation DES, complex lesions still have been associated with worse clinical outcomes (20). A randomized trial with 2,292 patients who received zotarolimus- or everolimus-eluting stents revealed that compared with patients with simple lesions, those with complex lesions had higher rates of target-lesion failure (6.3% vs. 9.3%, respectively; p = 0.015) and a patient-oriented composite endpoint at 1 year post-implantation (11.6% vs. 16.1%; p = 0.004) (20). An analysis of 8,061 everolimus-eluting stent-treated patients by Naidu et al. (21) revealed that total stent length was an independent predictor of stent thrombosis (HR per 10 mm: 1.30; p < 0.001). According to National Cardiovascular Data Registry results, patients with chronic total occlusions also had higher rates of MACEs compared with patients with nonchronic total occlusions (1.6% vs. 0.8%; p < 0.001) (22). The event rates of stent thrombosis and TLR in the present meta-analysis (0.5% and 5.3% in the angiographic-guidance arm, respectively) were generally similar to those in the previous randomized clinical trial using everolimus-eluting stents (1.0% and 2.4%, respectively) (23), and the clinical benefit in terms of MACEs in the present meta-analysis could be also attributed to the increased power from the use of randomized clinical trial and individual-level data from 2,345 patients. Patient-level meta-analyses compensate for limitations of study-level meta-analyses. This approach resulted in improved internal validity and allowed time-to-event and subgroup comparisons.
First, the identified randomized trials enrolled only patients with complex coronary lesions. Accordingly, external generalizability should be considered limited. Second, the 1-year follow-up period may not be sufficient for the assessment of long-term clinical outcomes.
Compared with angiographic guidance, IVUS-guided new-generation DES implantation was associated with favorable outcomes for the composite of cardiac death, myocardial infarction, or stent thrombosis, for complex coronary lesions.
WHAT IS KNOWN? Previous meta-analyses have indicated that IVUS-guided DES implantation was associated with a significant reduction in MACEs. These studies included patients who received first-generation DES and used only study-level (not patient-level) data. Published randomized trials that compare IVUS-guided versus angiography-guided new-generation DES implantation are scarce.
WHAT IS NEW? This is the first meta-analysis to use individual patient–level data in new-generation DES–treated patients, and the use of IVUS had clinical benefits for the hard clinical endpoint, the composite of cardiac death, myocardial infarction, and stent thrombosis.
WHAT IS NEXT? An individual-level meta-analysis with a greater number of randomized studies and a larger number of patients is needed to confirm these findings.
For supplemental data and figures, please see the online version of this article.
This study was supported by a grant from the Korea Healthcare Technology Research and Development Project, Ministry for Health and Welfare, South Korea (Nos. A085136 and HI15C1277), the Mid-Career Researcher Program through an NRF grant funded by the MEST, South Korea (No. 2015R1A2A2A01002731), and the Cardiovascular Research Center (Seoul, South Korea). The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- drug-eluting stent(s)
- hazard ratio
- intravascular ultrasound
- major adverse cardiac event(s)
- target lesion revascularization
- Received June 21, 2016.
- Revision received July 11, 2016.
- Accepted July 14, 2016.
- American College of Cardiology Foundation
- Jang J.S.,
- Song Y.J.,
- Kang W.,
- et al.
- Ahn J.M.,
- Kang S.J.,
- Yoon S.H.,
- et al.
- Elgendy I.Y.,
- Mahmoud A.N.,
- Elgendy A.Y.,
- Bavry A.A.
- Dangas G.D.,
- Serruys P.W.,
- Kereiakes D.J.,
- et al.
- Kim J.S.,
- Kang T.S.,
- Mintz G.S.,
- et al.
- Liberati A.,
- Altman D.G.,
- Tetzlaff J.,
- et al.
- Higgins J.P.,
- Altman D.G.,
- Gotzsche P.C.,
- et al.
- Kim B.K.,
- Shin D.H.,
- Hong M.K.,
- et al.
- Cutlip D.E.,
- Windecker S.,
- Mehran R.,
- et al.
- Thygesen K.,
- Alpert J.S.,
- Jaffe A.S.,
- et al.
- Navarese E.P.,
- Tandjung K.,
- Claessen B.,
- et al.
- Palmerini T.,
- Biondi-Zoccai G.,
- Della Riva D.,
- et al.
- Stefanini G.G.,
- Serruys P.W.,
- Silber S.,
- et al.
- Naidu S.S.,
- Krucoff M.W.,
- Rutledge D.R.,
- et al.
- Brilakis E.S.,
- Banerjee S.,
- Karmpaliotis D.,
- et al.