There are conflicting data regarding the benefit of intravascular ultrasound (IVUS)–guided percutaneous coronary intervention (PCI) over angiography-guided PCI. Since the last meta-analysis was published, several new studies have been reported. We performed a comprehensive meta-analysis to evaluate the clinical impact of IVUS-guided PCI with drug-eluting stent compared with conventional angiography-guided PCI. This meta-analysis included 26,503 patients from 3 randomized and 14 observational studies; 12,499 patients underwent IVUS-guided PCI and 14,004 underwent angiography-guided PCI. Main outcome measures were total mortality, myocardial infarction (MI), stent thrombosis, and target lesion revascularization (TLR). IVUS-guided PCI was significantly associated with more stents, longer stents, and larger stents. Regarding clinical outcomes, IVUS-guided PCI was associated with a significantly lower risk of TLR (odds ratio [OR] 0.81, 95% confidence interval [CI] 0.66 to 1.00, p = 0.046). In addition, the risk of death (OR 0.61, 95% CI 0.48 to 0.79, p <0.001), MI (OR 0.57, 95% CI 0.44 to 0.75, p <0.001), and stent thrombosis (OR 0.59, 95% CI 0.47 to 0.75, p <0.001) were also decreased. In conclusion, our meta-analysis demonstrated that IVUS-guided PCI was associated with lower risk of death, MI, TLR, and stent thrombosis after drug-eluting stent implantation.
Intravascular ultrasound (IVUS) has provided valuable information on cross-sectional coronary vascular structure and has played a key role in contemporary stent-based percutaneous coronary interventions (PCIs) by accurately assessing coronary anatomy, assisting in selection of treatment strategy, and defining optimal stenting outcomes. In the bare-metal stent era, randomized trials and meta-analysis demonstrated that IVUS-guided PCI was mainly associated with a lower risk of angiographic restenosis and target vessel revascularization (TVR). In the drug-eluting stent (DES) era, a recent meta-analysis showed that the risk of death and stent thrombosis (ST) was reduced by IVUS-guided DES implantation. Recently, 2 randomized trials and several observational studies have been reported since the last meta-analysis was published. Accordingly, we performed an updated comprehensive meta-analysis to evaluate the clinical impact of IVUS-guided PCI with DES implantation (compared with conventional angiography-guided PCI).
Methods
The literature was searched for studies that compared the clinical outcomes of IVUS-guided PCI with those of angiography-guided PCI and conformed to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement for conducting and reporting systematic reviews. A computerized search was performed of MEDLINE, EMBASE, and Cochrane databases from January 1995 to May 2013. Combinations of the following terms were used in the search process: “ultrasound, intravascular,” “IVUS,” “IVUS-guided,” “angiography-guided,” “percutaneous coronary intervention,” “PCI,” “drug-eluting stent,” and “stent.” Additional data sources included conference proceedings from major meetings of the American Heart Association, American College of Cardiology, and Transcatheter Cardiovascular Therapeutics. Reference lists of selected reports were reviewed for other potentially relevant citations. In the case of duplicate reports from the same patients, the most complete data were retrieved from the studies for quantitative synthesis. Additional searches for potential studies included references of review reports and earlier meta-analyses. Two investigators (J-MA and S-HY) independently screened the titles and abstracts and eventually examined the full texts of the original reports included in the study. Discrepancies were resolved by consensus.
Studies were included in the meta-analysis if they met the following prespecified criteria: (1) clinical studies published in peer-reviewed journals with fully available text, (2) studies comparing cohorts of IVUS-guided PCI with angiography-guided PCI with DES, (3) documentation of clinical outcomes of death, myocardial infarction (MI), repeat revascularization, or ST, separately or in combination, and (4) follow-up duration of ≥9 months. Reports of mixed treatment with bare-metal stent and DES implantation were excluded.
Patient characteristics, study design, and outcomes were systematically reviewed and recorded. The following outcomes were extracted: the number and length of implanted stents, minimal lumen diameter, mean stent diameter, and major adverse cardiac events (MACEs: defined as the composite of death, MI, and repeat revascularization) or individual outcomes of death, MI, target lesion revascularization (TLR), TVR, and ST. The quality of the retrieved studies was assessed to ensure minimization of bias, but no formal scoring system was used. Reviewers were not blinded to reports, publication sites, or author affiliations.
A random-effects meta-analysis was used to obtain the overall effect for the odds ratios (ORs) and the standardized mean difference for binary or continuous data. The DerSimonian and Laird model was used for the random-effects meta-analysis to account for excess variability (heterogeneity) across studies. Statistical heterogeneity across studies was assessed with the I 2 statistic that is derived from Cochran’s Q. I 2 values >25%, >50%, and >75% were considered as evidence of low, moderate, and severe statistical heterogeneity, respectively.
To examine the possible sources of heterogeneities, a random-effects meta-regression was performed to test whether any covariate was associated with an observed effect size. Publication bias (i.e., the likelihood of a small yet nominally significant study being selectively published) was examined by visual inspection of constructed funnel plots that relate the effect size to the precision of the effect estimate. Presence of asymmetry suggested possible publication bias. Egger’s test was employed to quantify the asymmetry.
A 2-tailed p value of <0.05 was considered statistically significant. All data analyses were performed with Comprehensive Meta-Analysis, version 2 (Biostat, Inc., New Jersey).
Results
The electronic search yielded 589 citations that were screened by reviewing the title or abstract. Of these, 33 publications were reviewed in full and 17 studies (26,503 patients) were included in the meta-analysis ( Figure 1 ). Of these patients, 12,499 underwent IVUS-guided PCI and 14,004 underwent angiography-guided PCI. Three studies were prospective randomized controlled studies, and the other 14 studies were prospective or retrospective observational studies.
Baseline characteristics of the included studies, major enrollment criteria, and the definition of clinical outcomes are provided in Table 1 . Two studies were dedicated to left main stenosis, 3 studies to bifurcation, 2 studies to long lesion, and 5 studies to real-world PCI population with minimal exclusion criteria. Overall, the average patient age was 63 years; 69% of patients were men, 31% had diabetes, and the average follow-up length was 29 months.
| Study | Publication Year | Enrollment Period | Patient Number | Design | Adjustment | Study Region | Follow-Up (Mo) | Age (yrs) | Men (%) † | Diabetes (%) | Hypertension (%) | Hyperlipidemia (%) | LVEF (%) | ACS (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Agostoni et al | 2005 | 2002–2003 | 24/34 | Observational | — | EU | 12 | 62/64 | 62/73 | 37/29 | 58/59 | 62/68 | 52/44 | 33/32 |
| Roy et al | 2008 | 2003–2006 | 884/884 | Observational | P | US | 12 | 66/66 | 69/70 | 36/34 | 82/82 | 86/87 | 47/48 | 76/75 |
| Park et al | 2009 | 2003–2006 | 145/145 | Observational | P | Asia | 36 | 64/65 | 70/70 | 34/34 | 59/59 | 29/30 | 61/63 | 63/61 |
| Jakabcin et al | 2010 | 2004–2005 | 105/105 | RCT | — | EU | 18 | 60/59 | 71/73 | 45/42 | 71/67 | 66/63 | — | 60/62 |
| Kim et al | 2010 | 2003–2006 | 303/112 | Observational | C | Asia | 48 | 60/61 | 73/72 | 23/21 | 47/55 | 22/24 | 59/59 | 48/51 |
| Claessen et al | 2011 | 2004–2006 | 631/873 | Observational | P | US/EU | 24 | 65/64 | 74/74 | 32/31 | 82/81 | 84/82 | — | 33/36 |
| Kim et al | 2011 | 2004–2006 | 487/487 | Observational | P | Asia | 36 | 62/62 | 67/67 | 32/33 | 60/58 | 35/35 | 61/59 | 53/56 |
| Youn et al | 2011 | 2003–2008 | 125/216 | Observational | — | Asia | 36 | 60/61 | 74/63 ∗ | 27/33 | 50/51 | 22/11 ∗ | 45/48 ∗ | 100/100 |
| Park et al | 2012 | — | 619/802 | Observational | P | Asia | 12 | 62/63 ∗ | 64/65 | 38/39 | 71/76 ∗ | 76/76 | — | 49/54 ∗ |
| Ahn et al | 2013 | 2008–2009 | 49/36 | Observational | — | Asia | 24 | 65/65 | 61/61 | 27/30 | 51/56 | 29/25 | 54/56 | 47/47 |
| Ahn et al | 2013 | 2008–2010 | 1,616/1,628 | Observational | P | Asia | 24 | 62/64 ∗ | 69/64 ∗ | 31/32 | 57/63 ∗ | 40/34 ∗ | 60/59 | 51/68 ∗ |
| Chen et al | 2013 | 2007–2010 | 324/304 | Observational | P | Asia | 12 | 63/65 | 81/75 | 19/18 | 67/61 | 33/35 | 61/60 | 87/79 ∗ |
| Chieffo et al | 2013 | — | 142/142 | RCT | EU | 24 | 64/64 | 82/77 | 24/27 | 70/67 | 70/77 | 55/56 | 30/26 | |
| Hur et al | 2013 | 2003–2006 | 2,765/1,816 | Observational | C | Asia | 36 | 60/63 ∗ | 71/67 ∗ | 26/27 | 47/51 ∗ | 23/20 ∗ | 58/55 ∗ | 54/59 ∗ |
| Kim et al | 2013 | — | 269/274 | RCT | Asia | 12 | 63/64 | 66/55 ∗ | 32/30 | 61/66 | 61/62 | 55/54 | 38/39 | |
| Yoon et al | 2013 | — | 662/912 | Observational | C | Asia | 12 | 61/63 ∗ | 65/65 | 28/30 | 60/63 | 61/56 | — | 60/55 |
| Witzenbichler et al | 2014 | — | 3,349/5,234 | Observational | C | US/EU | 12 | 63/64 ∗ | 73/75 | 31/33 | 78/81 ∗ | 68/78 ∗ | — | — |
Ten studies reported minimal lumen diameter by quantitative coronary angiographic analysis, and 12 studies reported mean stent diameter. Compared with angiography-guided PCI, IVUS-guided PCI was associated with larger stents and a larger postprocedure angiographic minimal lumen diameter. The mean difference in stent size and in postprocedural minimal lumen diameter was 0.33 mm (95% confidence interval [CI] 0.22 to 0.44, p <0.001) and 0.34 mm (95% CI 0.27 to 0.40, p <0.001), respectively. IVUS guidance was also associated with more stents and longer stents as reported in 9 studies and 13 studies, respectively. The mean difference in the number and length of stents used was 0.27 (95% CI 0.11 to 0.43, p <0.001) and 0.18 mm (95% CI 0.08 to 0.27, p <0.001), respectively ( Figure 2 ). Significant heterogeneity was observed across the studies regarding mean stent diameter, implanted stent number, and implanted stent length.
Periprocedural MI was reported in 6 studies. Of note, the risk of periprocedural MI did not significantly differ between IVUS-guided and angiography-guided DES implantation (OR 1.01, 95% CI 0.73 to 1.67, p = 0.65). However, severe heterogeneity was observed across the studies (I 2 = 81).
Long-term MACE was reported in 16 studies. IVUS-guided DES implantation was associated with a significant reduction of MACE (OR 0.74, 95% CI 0.64 to 0.85, p <0.001) with mild heterogeneity across the studies (I 2 = 46.2; Figure 3 ). Mortality was reported in 16 studies; IVUS-guided DES implantation resulted in a significant reduction of death (OR 0.61, 95% CI 0.48 to 0.79, p <0.001) with mild heterogeneity across the studies (I 2 = 42; Figure 3 ). MI was reported in 16 studies; IVUS-guided DES implantation led to a significant reduction of MI (OR 0.57, 95% CI 0.44 to 0.75, p <0.001) with mild heterogeneity across the studies (I 2 = 35; Figure 3 ). TVR was reported in 12 studies; the risk of TVR significantly reduced by IVUS-guided PCI (OR 0.82, 95% CI 0.70 to 0.97, p = 0.022) with mild heterogeneity across the studies (I 2 = 38.5; Figure 3 ). TLR was reported in 12 studies; the risk of TLR was significantly reduced by IVUS-guided PCI (OR 0.81, 95% CI 0.66 to 1.00, p = 0.046) with mild heterogeneity across the studies (I 2 = 41; Figure 3 ). ST was reported in 16 studies; IVUS-guided DES implantation was associated with a significant reduction of ST (OR 0.59, 95% CI 0.47 to 0.75, p <0.001) with no heterogeneity across the studies (I 2 = 5; Figure 3 ).
