There have been mixed results regarding the efficacy and safety of various percutaneous coronary intervention bifurcation techniques. An electronic search of Medline, Scopus, and Cochrane databases was performed for randomized controlled trials that compared the outcomes of any bifurcation techniques. We conducted a pairwise meta-analysis comparing the 1-stent versus 2-stent bifurcation approach, and a network meta-analysis comparing the different bifurcation techniques. The primary outcome was major adverse cardiac events (MACEs).
The analysis included 22 randomized trials with 6,359 patients. At a weighted follow-up of 25.9 months, there was no difference in MACE between 1-stent versus 2-stent approaches (risk ratio [RR] 1.20, 95% confidence interval [CI] 0.92 to 1.56). Exploratory analysis suggested a higher risk of MACE with a 1-stent approach in studies using second-generation drug-eluting stents, if side branch lesion length ≥10 mm, and when final kissing balloon was used. There was no difference between 1-stent versus 2-stent approaches in all-cause mortality (RR 0.95, 95% CI 0.69 to 1.30), cardiovascular mortality (RR 1.07, 95% CI 0.68 to 1.68), target vessel revascularization (TVR) (RR 1.22, 95% CI 0.90 to 1.65), myocardial infarction (MI) (RR 1.04, 95% CI 0.69 to 1.56) or stent thrombosis (RR 1.10, 95% CI 0.68 to 1.78). Network meta-analysis demonstrated that double kissing crush technique was associated with lower MACE, MI, TVR, and target lesion revascularization, whereas culotte technique was associated with higher rates of stent thrombosis. In this meta-analysis of randomized trials, we found no difference between 1-stent versus 2-stent bifurcation percutaneous coronary intervention approaches in the risk of MACE during long-term follow-up. Among the various bifurcation techniques, double kissing crush technique was associated with lower rates of MACE, target lesion revascularization, TVR, and MI.
Coronary bifurcation lesions account for ∼20% of epicardial coronary stenoses. Percutaneous coronary intervention (PCI) for these lesions remains challenging and has been linked with worse outcomes, compared with non-bifurcation lesions. Earlier randomized trials have demonstrated the superiority of a provisional or 1-stent strategy compared with a 2-stent strategy; however, there have been numerous advances in PCI techniques for bifurcation lesions since these trials. Recent randomized trials have demonstrated that a double kissing (DK) crush technique reduces the risk of adverse events in the subset of patients with left main true bifurcation lesion (involves significant [>50%] stenosis in both the main and side branches). , Recent guideline recommendations have endorsed that the DK-crush technique may be preferred in left main true bifurcation lesions (class IIb recommendation). A recent randomized trial has demonstrated that a 2-stent strategy reduced the composite of cardiac death, target vessel myocardial infarction (MI), or clinically driven target lesion revascularization (TLR) compared with provisional stenting (PS). The objective of this systematic review and meta-analysis was to pool data from randomized trials to compare 1-stent versus 2-stent approaches and to provide an indirect comparison for the efficacy and safety of various bifurcation techniques.
Methods
A computerized search of Medline, Scopus, and Cochrane databases was performed without language restriction through January 2021, using the terms “percutaneous coronary intervention,” “PCI” and “bifurcation” separately and in combination to identify randomized clinical trials that evaluated the outcomes of various bifurcation PCI techniques. We screened the bibliographies of the retrieved studies and ClinicalTrials.gov to identify any relevant studies not retrieved through the initial search. This meta-analysis was performed following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. A protocol for this meta-analysis was prospectively registered with PROSPERO (CRD42021233149).
We included randomized controlled clinical trials that compared the clinical outcomes of 2 or more bifurcation PCI techniques performed for chronic coronary syndromes (i.e., stable angina or silent ischemia) or acute coronary syndrome (ACS). For studies with multiple reports, we used data from the longest reported follow-up. We excluded trials including bifurcation PCI in cases with chronic total occlusion. The bifurcation PCI techniques evaluated in this systematic review included: PS, T-stenting/T and protrusion (T/TAP), crush, culotte, and DK-crush.
Two independent investigators (MS and AS) extracted the study design, baseline characteristics, intervention strategies, and clinical outcomes. Discrepancies in investigators were resolved by consensus.
The primary outcome of the study was the composite of major adverse cardiovascular events (MACEs). The secondary outcomes included all-cause mortality, cardiac mortality, MI, target vessel revascularization (TVR), TLR, and definite or probable stent thrombosis. Definition of end points was adopted as per each study protocol. Stent thrombosis events were defined per the Academic Research Consortium criteria.
The quality of the included trials was assessed using the Cochrane risk assessment tool of bias. The assessment tool included 7 criteria; random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other sources of bias. Accordingly, studies were classified into low risk, unclear risk, or high risk of bias.
Pairwise meta-analysis was used as the main analyses model, to compare 1-stent versus 2-stent bifurcation approaches. Network meta-analysis was conducted when comparing the 5 bifurcation PCI techniques (PS, T/TAP, crush, culotte, and DK-crush). Analysis of the clinical outcomes was performed by intention-to-treat . Data were pooled using random-effects models to account for the heterogeneity between the trials. Inconsistency was examined by comparing the deviance residuals and deviance information criterion statistics in fitted consistency and inconsistency models from the entire network on each node. Summary estimates were reported as risk ratios (RRs) and corresponding 95% confidence intervals (CIs). Statistical heterogeneity across trials was assessed by I 2 statistics, with I 2 statistic values <25%, 25% to 50%, and >50% considered as low, moderate, and a high degree of heterogeneity, respectively. , Publication bias was assessed using Egger’s test. Sensitivity analyses were conducted after excluding studies with a high risk of bias and studies with arms including multiple bifurcation PCI techniques. Meta-regression analyses were pre-specified for the primary outcome according to diabetes mellitus, proportion of patients with ACS, Medina classification (1,1,1), and the use of final kissing balloon (FKB). Exploratory sensitivity and meta-regression analyses were conducted for the primary outcome of MACE in the pairwise meta-analysis. P values were 2-tailed and considered statistically significant if <0.1 when evaluating subgroup interaction, and ≤0.05 in all other analyses. All analyses were conducted using the Rstudio software using “netmeta” and “meta” packages (RStudio Team (2020). RStudio: Integrated Development for R. RStudio, PBC, Boston, Massachusetts.)
Results
The study selection process is outlined in Figure 1 . The final analysis included 22 randomized studies with a total of 6,359 patients. , , , The included studies analyzed the following bifurcation techniques: PS (n = 2,171), T/TAP (n = 497), crush (n = 905), culotte (n = 873), DK-crush (n = 851) and group with mixed 2-stent bifurcation techniques (n = 1,062) ( Supplementary Table 1 ). The weighted mean follow-up was 25.9 months. The characteristics of the included studies are outlined in Table 1 . The bifurcation techniques evaluated included the proportion of presentations with ACS varied across studies. Studies in which the majority of included patients presented with ACS included Pan et al, DK-Crush I, Ye Feit et al (2010 and 2012), DK-Crush II, DK-Crush III, DK-Crush V, Zhang et al, Zheng et al and DEFINITION II trial. , , , , , , , The use of intracoronary imaging was reported in only a few studies, which were mostly the contemporary studies. There was considerable variation in the proportion of patients receiving intracoronary imaging across different studies (27.7% to 100%) There were relatively higher rates of crossover from 1-stent to 2-stent approach, up to ≈50% in some studies , ( Table 1 ).
| Study | No of centers | Interventions (group 1/ group 2) | Sample size (group 1/ group 2) | DES generation | Follow-up duration (m) | Mean duration of DAPT | ACS% | Intracoronary imaging use% | Cross over |
|---|---|---|---|---|---|---|---|---|---|
| Pan et al. 2004 | Multi-centered | PS vs T/TAP | 47/44 | First | 6 | 12m | 88 | NA | 2% crossed from PS to T/TAP 9% crossed from T/TAP to PS |
| Colombo et al. 2004 | Multi-centered | T/TAP vs PS | 63/23 | First | 6 | 3m | 17 | 100 | 4.7% crossed from T/TAP to PS 51.2% crossed from PS to T/TAP |
| DK-Crush I 2008 | Multi-centered | Crush vs DK Crush | 156/155 | First | 8 | 12 m | 85.8 | N/R | 0% crossed from Crush to DK Crush 0% crossed from DK Crush to Crush |
| CACTUS 2009 | Multi-centered | Crush vs PS | 177/173 | First | 6 | 6m | 46 | NA | 31% crossed from PS to Crush 0% crossed from Crush to PS |
| Ye Fei et al. 2010 | Single-centered | DK vs PS | 25/26 | First | 8 | ≥ 12 m | 86 | N/R | 0% crossed from Crush to DK Crush 0% crossed from DK Crush to Crush |
| Lin et al. 2010 | Single-centered | PS vs Two-stent | 54/ 54 | First | 8 | 12 m | 42 | N/R | 16.7% crossed from PS to Two-stent 5.6% crossed from Two-stent to PS |
| Hildick-Smith et al. 2010 | Multi-centered | PS vs Two-stent | 250/250 | First | 9 | ≥9 m | 34 | N/R | 2.8% crossed from PS to Two-stent 3.6% crossed from Two-stent to PS |
| Ye Fei et al. 2012 | Single-centered | DK crush vs PS | 38/30 | First | 12 | 12 m | 68 | N/R | 0% crossed from DK crush to PS 0% crossed from PS to DK crush |
| NSTS 2013 | Single-centered | Crush vs Culotte | 209/215 | First | 36 | 6 to 12 m | 22.87 | N/R | 0% crossed from Crush to Culotte 0% crossed from Culotte to Crush |
| NBS 2013 | Multi-centered | PS vs Two-stent | 202/202 | First | 60 | 6 to 12 m | 16 | N/R | 0% crossed from PS to Two-stent 0% crossed from Two-stent to PS |
| Ruiz et al. 2013 | Single-centered | PS/T-stenting | 33/36 | Second | 9 | 12 m | N/R | N/R | 9.1% crossed from PS to Two-stent 25% crossed from Two-stent to PS |
| Dk-Crush III 2015 | Multi-centered | DK crush/culotte | 210/209 | Second | 36 | 12 m | 90.2 | 71.4 | 0% crossed from Dk-crush to Culotte 0% crossed from Culotte to DK-crush |
| BBK I 2015 | Single-centered | PS/T-stenting | 101/101 | First | 60 | 6 m | 0 | N/R | 18.8% crossed from PS to T-stenting 3.0% crossed from T-stenting to PS |
| PERFECT 2015 | Multi-centered | Crush/PS | 213/206 | First and Second | 12 | 12 m | 38.4 | 95.70 | 0.5% crossed from Crush to PS 25.9% crossed from PS to Crush |
| NBBSIV 2015 | Multi-centered | PS/Two-stent | 218/228 | First and Second | 24 | 12 m | 14.8 | N/R | 3.7% crossed from PS to Two-stent 4% crossed from Two-stent to PS |
| Zhang et al. 2016 | Single-centered | PS/Culotte | 52/52 | First and Second | 9 | 12 m | 66.4 | N/R | 3.8% crossed from PS to Two-stent 0% crossed from Two-stent to PS |
| Hildick-Smith et al. 2016 | Multi-centered | PS/Culotte | 103/97 | Second | 12 | 12 m | 31.50 | N/R | 16% crossed from PS to Two-stent (T-stent) 2% crossed from Two-stent to PS |
| BBK II 2016 | Single-centered | Culotte/ TAP | 150/150 | First and Second | 12 | Clopidogrel (6 months)/ prasugrel or ticagrelor (12 months) | 20 | N/R | 0.7% crossed from TAP to Culotte 0% crossed from Culotte to TAP |
| Zheng et al. 2016 | Single-centered | Crush/Culotte | 150/150 | NA | 12 | 12 m | 91.4 | N/R | N/R |
| DK-Crush II 2017 | Multi-centered | DK crush / PS | 183/183 | First | 60 | 12 m | 84 | 47 | 28% crossed from PS to Two-stent 0% crossed from Two-stent to PS |
| Dk-Crush V 2019 | Multi-centered | PS/ DK crush | 242/240 | Second | 36 | 12 m | 72.2 | 41 | 47.1% crossed from PS to Two-stent 0% crossed from Two-stent to PS |
| DEFINITION II 2020 | Multi-centered | Two-stent / PS | 328/325 | First and Second | 12 | 12 m | 72 | 27.70 | 22.5% crossed from PS to Two-stent 7.9% crossed from Two-stent to PS |
The baseline and procedural characteristics for patients in the included studies appear in Tables 2 and 3 . The reported mean left ventricular ejection fraction for included patients was >50% in all studies. Most studies evaluated patients predominantly with true bifurcation lesions, and the Medina classification for bifurcation lesions was more commonly 1,1,1. The DK-crush III and DK-crush V trials exclusively included patients with bifurcation PCI involving left main coronary artery (LMCA), , whereas the proportion of LMCA bifurcation PCI was lower in other studies. Few studies reported using radial access for bifurcation PCI, with the highest proportion of radial access in the DEFINITION II trial (78.7% and 80.6% in PS and 2-stent groups, respectively). Procedural success >90% was achieved in all studies except Colombo et al , in which the PS arm had only 77.3% procedural success.
| Study | Age mean(+- SD) (group 1/group2) | Male % (group1/group2) | HTN % (group1/group2) | DM % (group1/group2) | prior MI % (group1/group2) | Prior PCI % (group1/group2) | LVEF mean+/-SD (group1/group2) | FH of CAD% (group1/group2) | HLD% (group1/group2) | Tobacco use % (group1/group2) |
|---|---|---|---|---|---|---|---|---|---|---|
| Pan et al. 2004 | 61±10/ 58±11 | 72/86 | 59/57 | 42/39 | 19/39 | NA | 60± 11/ 55± 11 | NA | 53/41 | 38/52 |
| Colombo et al. 2004 | 63±10/ 62±9 | 76/91 | N/R | 21/26 | N/R | N/R | 59± 10/ 59± 9 | N/R | N/R | N/R |
| DK-Crush I 2008 | 64±9 / 64±9 | 70 / 76 | 77/ 76 | 8/27 | 12/9 | 11/12 | 63± 13/ 62± 11 | N/R | 63 %/ 69 % | 63/ 64 |
| CACTUS 2009 | 65±10/ 67±10 | 80 / 76 | 71/ 80 | 24 / 22 | 45 / 35 | 31/27 | 55± 9/ 57± 8 | 47/36 | 64/71 | 20/17 |
| Ye Fei et al. 2010 | 63.6±11.5/ 63.2±9.9 | 64/73 | 76/73 | 16/19 | NA | NA | 59± 10/ 57± 10 | NA | 16/12 | NA |
| Lin et al. 2010 | 60.6±7.5/ 59.2±7.2 | 83/ 76 | 91/83 | 18.5/13 | 22.2/18.5 | 24/24 | 56± 6/ 57± 6 | NA | NA | 30/24 |
| Hildick-Smith et al. 2010 | 64±10/ 64±11 | 77/77 | 57/62 | 13/11 | 23/25 | 17/16 | 91% LVEF> 50 / 85% LVEF > 50 | 42/41 | 76/76 | 17/17 |
| Ye Fei et al. 2012 | 63.5±10.5/61.7±9.4 | 63/77 | 76/67 | 18/13 | 11/7 | NA | 62± 10/64± 6 | NA | 18/20 | NA |
| NSTS 2013 | 65±10/ 65±11 | 71/71 | 62/60 | 13/15 | NA | 40/34 | 57± 11 / 57± 12 | 57/62 | 84/74 | 20/27 |
| NBS 2013 | 63±10/ 63±10 | 76/78 | 54/58 | 13/12 | NA | 25/25 | N/R | 58/54 | 78/72 | NA |
| Ruiz et al. 2013 | 63.4±13.0/ 63.6 ±13.1 | 85/78 | 67/72 | 45/33 | N/R | 21/25 | N/R | N/R | 51/64 | 61/50 |
| Dk-Crush III 2015 | 64.3 ± 10.3/ 63.3 ± 9.2 | 77.1/79.9 | 70.5/61.2 | 31.9/30.1 | 15/14 | 22/15 | 59± 11/59 ± 11 | N/R | 41.4/42.1 | 58/54 |
| BBK I 2015 | 66.7± 9.2/66.9±10.5 | 79.4/78.2 | 92.1/89.1 | 25.7/18.8 | 18.8/20.8 | 44.6/51.5 | 59 ± 12/ 61 ± 12 | N/R | N/R | 9.9/13.9 |
| PERFECT 2015 | 60.9± 8.9/61.1±8.8 | 75.1/75.2 | 55.4/55.3 | 25.8/29.1 | 4.2/4.4 | 9.4/ 5.3 | 60.4 ± 6.8/59.5 ± 7.2 | 14.1/12.6 | 62.0/57.3 | 25.4/32.5 |
| NBBSIV 2015 | 64 ± 12/ 63 ± 11 | 74.7/78.2 | 70.0/65.6 | 16.5/15.4 | N/R | 35.5/ 33.5 | 57±6/56±7 | 50.5/ 47.4 | 82/81.1 | 18.9/21.1 |
| Zhang et al. 2016 | 64.52 ± 10.67/64.19 ± 7.27 | 92/83 | 67.3/63.5 | 19.2/21.2 | 23.1/19.2 | 25/23.1 | N/R | N/R | 11.5/ 11.5 | 59.6/51.9 |
| Hildick-Smith et al. 2016 | 62.9 (10.8)/63.5 (12.1) | 85/78 | 63/68 | 25/31 | 39/41 | 40/41 | N/R | 48/49 | 70/70 | 56/50 |
| BBK II 2016 | 66.3 ± 10.6/ 69.1 ± 10.3 | 71.3/76.0 | 88.0/85.3 | 27.3/28.0 | 16.0/21.3 | 38/32 | 56 ± 7.3 / 57 ± 6.0 | 40.7/39.3 | N/R | 11.3/11.3 |
| Zheng et al. 2016 | 63 ± 8/64 ± 9 | 72.7/74.0 | 70.7/72.7 | 22.0/24.7 | N/R | 26.7/22.7 | N/R | 30.0/34.7 | 76.0/70.0 | 38.7/44.7 |
| DK-Crush II 2017 | 63.9±11.1/64.7±10.0 | 78.8/75.8 | 65.2/60.9 | 19.6/23.1 | 17.4/14.2 | 21.2/20.9 | N/R | N/R | 33.7/29.1 | 31/24 |
| Dk-Crush V 2019 | 64 ± 10 /65 ± 9 | 77.7/82.9 | 64.5/72.9 | 25.6/28.8 | 21.1/21.7 | 18/ 14 | N/R | N/R | 47.5/47.5 | 32/34 |
| DEFINITION II 2020 | 63 ± 11/ 64 ± 10 | 77.7/76.9 | 66.2/70.1 | 34.1/35.7 | 11.9/12.9 | 19.8/16.6 | 59 ± 10/ 60 ± 10 | N/R | 69.2/68.6 | 28.4/30.2 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
