Controversies exist regarding the association between plaque composition and distal embolization phenomenon after percutaneous coronary intervention (PCI). We evaluated the effect of plaque characteristics on embolization after PCI by grayscale and virtual histology-intravascular ultrasound (IVUS). We searched PubMed, Ovid MEDLINE, and Cochrane databases for IVUS studies evaluating the coronary plaque characteristics in no reflow, distal embolization, and periprocedural myocardial infarction after PCI. Sixteen studies were included, totaling 1,697 patients who underwent PCI (292 patients with embolization and 1,405 patients without embolization). At the minimum lumen sites, the external elastic membrane (weighted mean difference 2.38 mm 2 , 95% confidence interval [CI] 1.02 to 3.74) and the plaque and media cross-sectional areas (weighted mean difference 2.44 mm 2 , 95% CI 1.44 to 3.45) were significantly greater in the embolization group than in the no embolization group. Pooled analysis showed that the absolute necrotic core volume (standardized mean difference 0.49, 95% CI 0.13 to 0.85), absolute (standardized mean difference 0.73, 95% CI 0.14 to 1.31) and relative (standardized mean difference 1.02, 95% CI 0.72 to 1.31) necrotic core areas at the minimum lumen sites were significantly greater in the embolization group than in the no embolization group, but the other plaque components were similar in the 2 groups. In conclusion, the necrotic core component derived from virtual histology-IVUS and the morphologic characteristics of plaque derived from grayscale IVUS are closely related to the distal embolization phenomenon after PCI.
Numerous studies have shown a relation between virtual histology-intravascular ultrasound (VH-IVUS) plaque composition and the no reflow phenomenon after percutaneous coronary intervention (PCI). Many IVUS studies have reported that the necrotic core component is the responsible plaque type for the occurrence of no reflow or distal embolization after PCI. However, other studies have demonstrated that fibrotic or fibrofatty plaque are prerequisites for the post-PCI no reflow phenomenon. We, therefore, performed a systematic review and meta-analysis of published studies evaluating characteristic lesion morphology and plaque composition observed by grayscale and/or VH-IVUS in patients with the no reflow phenomenon, distal embolization, and myocardial infarction after PCI.
Methods
We identified relevant studies through electronic searches of PubMed, Ovid MEDLINE, and the Cochrane Central Register of Controlled Trials from January 2001 through February 2012. Medical subject headings and keyword searches included the terms “intravascular ultrasound,” “grayscale,” “virtual-histology,” “coronary plaque,” “atherosclerotic plaque,” “no reflow,” “embolization,” “myocardial infarction,” “myonecrosis,” and “percutaneous coronary intervention.” The reference lists of the selected studies were systematically reviewed for other potentially relevant citations.
Two investigators (J.-S.J. and K.-I.C.) independently conducted the data search, data extraction, and quality assessment using a standardized approach. The selected publications were reviewed by the same investigators to determine whether the studies met the inclusion criteria: (1) grayscale and/or VH-IVUS examination before stenting, (2) a comparison of lesion characteristics and/or plaque composition in patients with or without no reflow/distal embolization phenomenon after primary PCI for acute myocardial infarction or after elective PCI or periprocedural myocardial infarction after elective PCI. Studies with inappropriate quantitative or qualitative IVUS measurements were excluded from the present study. The final inclusion of the studies was by agreement of both reviewers.
Two reviewers (J.-S.J. and Y.H.P.) extracted the relevant information from the studies, including patient characteristics (mean age, gender distribution, and risk factors), study period, publication year, sample size, and left ventricular ejection fraction. The reviewers were not unaware of the studies, publication sites, and investigator affiliations. If additional information was needed, the investigators were contacted.
Basically, we retrieved conventional IVUS data about the quantitative measurements, such as external elastic membrane (EEM) and plaque and media cross-sectional areas (CSAs), and plaque burden. Data regarding morphologic plaque features, such as eccentric plaque, plaque rupture, attenuated plaque, lipid pool-like image, and superficial/deep calcium were also collected, if available. Our primary end points of interest were the absolute and relative volumes of each plaque defined by VH-IVUS (e.g., fibrous, fibrofatty, dense calcium, and necrotic core). We also analyzed the absolute and relative areas of each plaque.
In the individual studies, the EEM and lumen CSAs were measured. The plaque and media CSA was calculated as the EEM minus the lumen CSA, and the plaque burden was calculated as the plaque and media CSA divided by the EEM CSA ×100. Coronary artery remodeling was assessed by comparing the lesion site to the reference segment EEM CSA. The VH-IVUS analysis classified the color-coded tissue into 4 major components: green (fibrotic), yellow-green (fibrofatty), white (dense calcium), and red (necrotic core). The results of the VH-IVUS analysis in each of the included studies were reported in absolute amounts and/or as a percentage of plaque area or volume.
For the comparison of the plaque characteristics between the 2 groups, fixed (Mantel-Haenszel method) or random effects models (Dersimonian and Laird method) were used to produce across-study weighted mean differences, standardized mean differences, and summary odds ratios (ORs) with 95% confidence intervals (CIs). All p values were 2-tailed, with statistical significance set at 0.05. To assess the effect of individual studies on the summary estimate of effect, we performed an exclusion sensitivity analysis, in which the pooled estimates were recalculated, omitting 1 study at a time.
We assessed the statistical heterogeneity between trials with the I 2 statistic, which is derived from Cochran’s Q [100 × (Q − df)/Q]. An I 2 value >25%, >50%, and >75% was considered evidence of low, moderate, and severe statistical heterogeneity, respectively. The likelihood of a publication bias was assessed graphically by generating a funnel plot for the remodeling index and mathematically using Egger’s test (p for significant asymmetry <0.1). For specific evaluation of the presence and extent of a publication bias, we used the trim-and-fill method according to Duval and Tweedie, which imputes missing studies in the funnel plot according to symmetry assumptions.
All statistical analyses were performed using the Review Manager, version 5.1 (Nordic Cochrane Center, Copenhagen, Denmark) and MIX, version 2.0 (BiostatXL, Sunnyvale, California).
Results
At the beginning, a total of 389 publications from January 2001 to February 2012 were screened. Of these, 354 publications were excluded because they did not meet the inclusion criteria. Of the remaining 35 complete reports, 19 studies without appropriate IVUS measurements or a comparison of IVUS findings were excluded. Finally, 16 studies were included in the quantitative synthesis ( Figure 1 ). Of these, 12 studies used the “no reflow phenomenon,” 3 studies used the “distal embolization phenomenon,” and 1 study used “periprocedural myonecrosis” as the embolization type.
Of the 1,697 patients in the final analysis, 292 developed 1 of the 3 types of embolization related to PCI. Of the 16 studies, 9 used both grayscale and VH-IVUS and 7 used grayscale IVUS only. The characteristics of the included studies and baseline demographics of the study populations are presented in Table 1 .
| Study | Year | Subjects (n) | Inclusion Criteria | IVUS Type | Age (yrs) | Men (%) | DM (%) | HT (%) | Hyperlipidemia (%) | Smoking (%) | LVEF (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Tanaka et al | 2002 | 13/87 | No reflow in AMI | Grayscale | 62/64 | 77/77 | 31/25 | 53/47 | 69/38 | 85/55 | NA |
| Watanabe et al | 2003 | 21/60 | No reflow in AMI | Grayscale | 65/65 | 70/76 | 25/29 | 42/62 | 33/43 | 57/57 | NA |
| Fukuda et al | 2003 | 12/128 | Distal embolization in AMI | Grayscale | 67/63 | 75/75 | 25/29 | 67/64 | 50/56 | 75/77 | NA |
| Iijima et al | 2006 | 20/200 | No reflow in ACS | Grayscale | 69/65 | 75/82 | 35/31 | 75/65 | 60/55 | NA | NA |
| Katayama et al | 2006 | 12/58 | No reflow in AMI | Grayscale | 65/63 | 75/76 | 25/34 | 58/60 | 42/48 | 75/71 | 51/51 |
| Nakamura et al | 2007 | 8/42 | No reflow in AMI | Grayscale and VH | 59/65 | 88/86 | 38/24 | 25/43 | 63/62 | 63/45 | NA |
| Kawaguchi et al | 2007 | 11/60 | Distal embolization in STEMI | Grayscale and VH | 64 | 70 | 28 | 45 | 29 | 33 | NA |
| Kawamoto et al | 2007 | 13/16 | Distal embolization in elective PCI | Grayscale and VH | NA | NA | NA | NA | NA | NA | NA |
| Bae et al | 2008 | 12/45 | No reflow in AMI | Grayscale and VH | 68/56 | 67/82 | 33/13 | 33/40 | 27/31 | 33/60 | 55/56 |
| Higashikuni et al | 2008 | 9/40 | No reflow in ACS | Grayscale and VH | 61/67 | 78/93 | 56/30 | 56/70 | 89/65 | 33/35 | 57/56 |
| Hong et al | 2009 | 38/42 | Myonecrosis in elective PCI | Grayscale and VH | 63/65 | 76/48 | 32/24 | 74/64 | NA | 40/43 | 59/59 |
| Ohshima et al | 2009 | 20/24 | No reflow in STEMI | Grayscale and VH | 74/66 | 65/83 | 60/54 | 75/58 | 45/63 | 40/67 | NA |
| Lee et al | 2009 | 30/263 | No reflow in ACS | Grayscale | 62/65 | 75/56 | 28/29 | 87/80 | 82/78 | 40/35 | NA |
| Endo et al | 2010 | 30/140 | No reflow in STEMI | Grayscale | 65/63 | 70/81 | 20/26 | 53/61 | 57/52 | 67/71 | NA |
| Hong et al | 2011 | 24/166 | No reflow in ACS | Grayscale and VH | 60/61 | 58/66 | 25/19 | 71/52 | NA | 54/49 | 62/62 |
| Ohshima et al | 2011 | 19/34 | No reflow in STEMI | Grayscale and VH | 73/67 | 68/85 | 26/38 | 79/71 | 47/62 | 37/47 | NA |
At the minimum lumen sites, the EEM CSA was significantly greater in the embolization group than in the no embolization group (weighted mean difference 2.38 mm 2 , 95% CI 1.02 to 3.74, p <0.001). Evidence of moderate heterogeneity was seen among the included studies (heterogeneity chi-square = 19.04, I 2 = 58%, p = 0.01). Similarly, the plaque and media CSA at the site of minimal lumen area was also significantly greater in the embolization group (weighted mean difference 2.44 mm 2 , 95% CI 1.44 to 3.45, p <0.001) with evidence of low statistical heterogeneity (heterogeneity chi-square = 6.84, I 2 = 27%, p = 0.23). Accordingly, the plaque burden was significantly greater in the embolization group with an weighted mean difference of 4.0 % (95% CI 1.73 to 6.27, p <0.001), with mild heterogeneity (heterogeneity chi-square = 16.68, I 2 = 58%, p = 0.02). The remodeling index was reported in 9 studies. It was significantly greater in the embolization group compared to the no embolization group (weighted mean difference 0.08, 95% CI 0.04 to 0.11; Figure 2 ) with no evidence of statistical heterogeneity (heterogeneity chi-square = 5.53, I 2 = 0%, p = 0.70). Positive remodeling was also more frequent in the embolization group (OR 1.75, 95% CI 1.17 to 2.63, p = 0.007).
In terms of the grayscale morphologic features of culprit plaque known to be associated with no reflow phenomenon or distal embolization after PCI, eccentric plaque (OR 2.76, 95% CI 1.42 to 5.40), ruptured plaque (OR 4.51, 95% CI 2.59 to 7.86), and attenuated plaque (OR 8.30, 95% CI 2.85 to 24.15) were more frequently found in the embolization group than in the no embolization group ( Supplementary Figure 1 ). The prevalence of lipid pool-like image tended to be greater in the embolization group. However, the incidence of superficial and deep calcium was not different between the 2 groups.
Six studies were used for the analysis of the absolute volumes of each plaque. Although no significant differences were found in the fibrous, fibrofatty, and dense calcium volumes between the 2 groups, the absolute necrotic core volume was significantly greater in the embolization group compared to no embolization group (standardized mean difference 0.49, 95% CI 0.13 to 0.85; Supplementary Figure 2 ). However, the relative volumes of each plaque were similar between the 2 groups ( Supplementary Figure 3 ). Because extensive heterogeneity was noted for relative plaque volumes across the included studies, we did a sensitivity analysis. After exclusion of the study by Nakamura et al, the fibrous plaque volume was significantly smaller in the embolization group (standardized mean difference −0.48, 95% CI −0.85 to −0.10, p = 0.001). However, the relative necrotic core volume (standardized mean difference 0.61, 95% CI 0.24 to 0.98, p = 0.001) and dense calcium volume (standardized mean difference 0.32, 95% CI 0.03 to 0.61, p = 0.03) were significantly greater in the embolization group compared with the no embolization group. However, none of the other studies influenced the summary estimates of relative plaque volumes to an extent that the conclusion would have been changed.
The funnel plot of the remodeling index showed the existence of a publication bias, confirmed using positive Egger’s regression-based test (p = 0.09). The results of the trim-and-fill method indicated that 4 missing studies were needed to achieve a symmetric funnel plot ( Figure 3 ).
