Percutaneous coronary intervention (PCI) of bifurcation lesions remains challenging with a higher risk of adverse outcomes. Whether adjunctive intravascular ultrasound (IVUS) imaging improves outcomes of PCI of bifurcation lesions remains unclear. This study sought to determine the long-term clinical outcomes associated with using IVUS for percutaneous treatment of coronary bifurcation lesions. From April 2003 through August 2010, 449 patients with 471 bifurcation lesions underwent PCI with (n = 247) and without (n = 202) the use of IVUS. Clinical outcomes (death, myocardial infarction [MI], periprocedural MI, stent thrombosis, target vessel revascularization [TVR], and target lesion revascularization [TLR]) were compared between patients undergoing PCI with and without IVUS using univariate and propensity score-adjusted analyses. Most patients (61%) presented with acute coronary syndrome and 89% of bifurcations lesions were Medina class 1,1,1. After propensity score adjustment, use of IVUS was associated with significantly lower rates of death or MI (odds ratio 0.38, 95% confidence interval 0.20 to 0.74, p = 0.005), death (odds ratio 0.40, 95% confidence interval 0.18 to 0.88, p = 0.02), MI (odds ratio 0.37, 95% confidence interval 0.14 to 0.98, p = 0.04), periprocedural MI (odds ratio 0.45, 95% confidence interval 0.20 to 0.97, p = 0.04), TVR (odds ratio 0.28, 95% confidence interval 0.14 to 0.53, p <0.0001), and TLR (odds ratio 0.27, 95% confidence interval 0.14 to 0.53, p = 0.0003) compared to no IVUS. In conclusion, IVUS-guided treatment of complex bifurcation lesions was associated with significantly lower rates of adverse cardiac events at late follow-up. Further study is warranted to evaluate the role of IVUS guidance in improving long-term outcomes after PCI of bifurcation lesions.
Percutaneous coronary intervention (PCI) of coronary bifurcation lesions represent 10% to 15% of PCIs and is technically challenging owing to multiple factors including anatomic variability related to bifurcation site, plaque burden and morphology, bifurcation angle, and branch diameter. Moreover, bifurcation anatomy may have dynamic variability during PCI, with plaque shift or dissection requiring adjustments in interventional approach. For any given bifurcation lesion, uncertainty exists in the selection of the most appropriate interventional strategy and of the potential benefit of available adjunctive technologies. Given its ability to more precisely define coronary anatomy, plaque characteristics, and stent dimensions and morphology compared to angiography, we hypothesized that intravascular ultrasound (IVUS) would improve the outcomes of PCI of bifurcation lesions. We compared long-term clinical outcomes of patients with coronary bifurcations lesions who underwent PCI after the availability of drug-eluting stents with and without the use of IVUS.
Methods
From April 2003 through August 2010, 9,131 patients underwent PCI at our institution. A retrospective analysis identified 482 patients with de novo coronary bifurcation lesions with a visually estimated diameter stenosis >50% involving the main branch at or within 5 mm of the origin of a side branch >1.5 mm in diameter. Patients were separated into groups by use of IVUS or no IVUS. Of the 482 patients, 13 patients (IVUS, n = 7; no IVUS, n = 6) had no clinical follow-up and could not be contacted by telephone and thus were excluded from the study. In addition, 20 patients were excluded owing to having multiple bifurcation lesions where IVUS was used during PCI on 1 bifurcation lesion but not used on another. The final analysis comprised 449 patients with 471 distinct coronary bifurcation lesions. All bifurcation lesions were classified according to the Medina classification. Baseline patient data were obtained through review of the electronic medical record and/or paper charts. Institutional review board approval was obtained on April 30, 2010 (protocol number 10-0475).
Each coronary angiogram was reviewed by an investigator independent of the operator. All patients were treated with aspirin 325 mg before the procedure. A loading dose of clopidogrel 300 to 600 mg or 60 mg prasugrel was administered before or at the time of the index procedure. All patients received anticoagulation during the procedure, but the choice of agent and dosage and the addition of glycoprotein IIb/IIIa inhibitors were left to the operator’s discretion. After PCI all patients were prescribed aspirin 81 to 325 mg/day indefinitely and clopidogrel 75 mg/day or prasugrel 10 mg/day for ≥12 months.
All PCIs were performed using standard techniques. The operator determined the stenting technique and type of stent(s) used to treat the bifurcation lesion. Final balloon inflation was performed in all cases. Decision to use IVUS was at the operator’s discretion at the time of PCI. All IVUS studies were performed using a commercially available system (20-MHz Eagle Eye gold or platinum IVUS catheter, Volcano Corp., Rancho Cordova, California; 40-MHz iCross IVUS catheter, Boston Scientific, Corp., Natick, Massachusetts) after intracoronary administration of nitroglycerin 100 to 200 μg. IVUS images were obtained using a manual or automated transducer pullback and were recorded on digital media. IVUS was performed in selected patients in the IVUS group before PCI to identify plaque burden and composition, vessel size(s), bifurcation angle, and/or mark side branch ostium. After stenting IVUS was used in all patients in the IVUS group to assess stent position, lesion coverage, degree of expansion, neocarina formation, and/or any complications of PCI not apparent by angiography (e.g., edge dissection, plaque shift, stent underexpansion).
The primary outcome of the study was a composite of death from any cause or myocardial infarction (MI). Secondary outcomes included individual end points of death, MI, periprocedural MI, target vessel revascularization (TVR), target lesion revascularization (TLR), or stent thrombosis. Patients were followed from the date of index stenting through December 31, 2010. All outcomes were adjudicated independently without knowledge of a patient’s IVUS status. MI, periprocedural MI (creatine kinase-MB increase >3 times the upper reference limit), TVR, and stent thrombosis were defined according to guidelines set forth by the Academic Research Consortium. TLR was defined as repeat revascularization with stenosis ≥50% within the stent, within 5-mm adjacent to the stent, or within 5-mm of the carina in the main or side branch. For patients with multiple coronary bifurcation lesions, death, MI, or periprocedural MI were counted only 1 time and analyzed per patient, whereas TVR, TLR, and stent thrombosis were analyzed per bifurcation lesion.
Differences between the IVUS and no-IVUS groups were analyzed by univariate methods. Categorical variables were presented as counts and percentages and compared using chi-square or Fisher’s exact tests. Continuous variables were presented as mean ± SD and compared using t tests or Mann–Whitney U tests. Patient-level outcomes (death or MI, death, MI, and periprocedural MI) were compared using Fisher’s exact test. For lesion-level outcomes (TLR, TVR, and stent thrombosis), generalized estimating equations were used to build logistic models.
Propensity scores were created for the IVUS versus no-IVUS groups based on patient and lesion characteristics and were used as an adjustment factor in a logistic regression model for each outcome. The following variables were used to calculate the propensity score: age, male gender, diabetes mellitus, hypertension, hyperlipidemia, current/former smoker, previous PCI, previous coronary artery bypass graft(s), peripheral arterial disease, chronic kidney disease, ST-segment elevation MI, non–ST-segment elevation MI, unstable angina, stable angina, asymptomatic/atypical chest pain, heparin, bivalirudin, glycoprotein IIb/IIa inhibitor, cilostazol, β blocker, angiotensin-converting enzyme/angiotensin receptor blocker, statin, bifurcation vessel, number of stents, stent technique, and stent type. Covariate balance between groups was evaluated by Wald chi-square statistic before and after propensity score adjustment. An additional analysis on matched propensity scores was also conducted. A greedy matching algorithm was used to match scores and standardized differences were calculated to determine covariate balance before and after matching. Logistic models were created using generalized estimating equation methods for all outcomes to account for matching. All statistical analyses were performed using SAS 9.3 (SAS Institute, Cary, North Carolina). A p value <0.05 was considered statistically significant.
Results
Baseline clinical characteristics for the 449 patients in the study are listed in Table 1 . Patients in the IVUS group were younger and had lower rates of hypertension, previous PCI, peripheral arterial disease, increased bivalirudin use, and decreased use of heparin or glycoprotein IIb/IIIa inhibitors compared to the no-IVUS group. Most patients (61%) presented with an acute coronary syndrome, where non–ST-segment elevation MI was more prevalent in the IVUS group and stable angina was more prevalent in the no-IVUS group. The 2 groups received similar pharmacologic treatments after PCI.
| Variable | IVUS | p Value | |
|---|---|---|---|
| Yes | No | ||
| (n = 247 patients) | (n = 202 patients) | ||
| Age (years) | 62.0 ± 11.9 | 65.9 ± 12.2 | 0.0006 |
| Men | 163 (66%) | 132 (65%) | 0.75 |
| Medical history | |||
| Diabetes mellitus | 76 (31%) | 77 (38%) | 0.89 |
| Hypertension | 198 (80%) | 180 (89%) | 0.001 |
| Hyperlipidemia | 188 (76%) | 166 (82%) | 0.12 |
| Current/former smoker | 132 (53%) | 107 (53%) | 0.92 |
| Previous percutaneous coronary intervention | 57 (23%) | 70 (35%) | 0.007 |
| Previous coronary artery bypass graft | 46 (19%) | 36 (18%) | 0.83 |
| Peripheral arterial disease | 17 (7%) | 25 (12%) | 0.05 |
| Chronic kidney disease | 30 (12%) | 24 (12%) | 0.93 |
| Diagnosis at percutaneous coronary intervention | |||
| ST-segment elevation myocardial infarction | 24 (10%) | 18 (9%) | 0.77 |
| Non–ST-segment elevation myocardial infarction | 78 (32%) | 44 (22%) | 0.02 |
| Unstable angina pectoris | 56 (23%) | 55 (27%) | 0.27 |
| Stable angina pectoris | 50 (20%) | 59 (29%) | 0.03 |
| Asymptomatic/atypical chest pain | 39 (16%) | 26 (13%) | 0.38 |
| Antithrombotic therapy during intervention | |||
| Heparin | 116 (47%) | 138 (68%) | <0.0001 |
| Bivalirudin | 119 (48%) | 61 (30%) | 0.0001 |
| Glycoprotein IIb/IIIa inhibitor | 121 (49%) | 123 (61%) | 0.01 |
| Discharge medications | |||
| Aspirin | 247 (100%) | 202 (100%) | |
| Clopidogrel | 242 (98%) | 202 (100%) | 0.06 |
| Prasugrel | 5 (2%) | 0 (0%) | 0.06 |
| Cilostazol | 5 (2%) | 3 (1%) | 0.74 |
| β Blocker | 220 (89%) | 183 (91%) | 0.60 |
| Angiotensin-converting enzyme/angiotensin receptor blocker | 184 (74%) | 149 (74%) | 0.86 |
| Statin | 232 (94%) | 193 (96%) | 0.45 |
Lesion characteristics and stenting techniques are listed in Table 2 . The IVUS group had significantly more lesions located at the left main bifurcation, whereas the no-IVUS group had more lesions located at the left circumflex/obtuse marginal artery and right coronary artery bifurcations. The predominant lesion type was Medina type 1,1,1 (89%). The most common PCI technique used in the 2 groups was provisional stenting of the main branch with angioplasty of the side branch, but this technique was employed more often in the IVUS group compared to the no-IVUS group. Most PCIs in the 2 groups included implantation of drug-eluting stents, but with a significantly higher frequency in the IVUS group compared to the no-IVUS group. Mean stent length was significantly longer in the IVUS group versus the no-IVUS group (17.1 ± 5.2 vs 15.0 ± 4.6 mm, p <0.0001). Significantly fewer stents were placed in the IVUS group compared to the no-IVUS group (1.6 ± 0.7 vs 1.8 ± 1.0, p = 0.007).
| Variable | IVUS | p Value | |
|---|---|---|---|
| Yes | No | ||
| (n = 258 lesions) | (n = 213 lesions) | ||
| Bifurcation coronary artery | |||
| Left main/left anterior descending/left circumflex coronary arteries | 37 (14%) | 4 (2%) | <0.0001 |
| Left anterior descending/diagonal coronary artery | 129 (50%) | 94 (44%) | 0.2 |
| Left circumflex/obtuse marginal artery | 56 (22%) | 70 (33%) | 0.007 |
| Right coronary artery/posterior descending artery/posterior left ventricular branch | 36 (14%) | 45 (21%) | 0.04 |
| Lesion classification | |||
| Medina classification | 0.69 | ||
| 0,0,1 | 5 (2%) | 2 (1%) | |
| 0,1,0 | 3 (1%) | 2 (1%) | |
| 0,1,1 | 6 (2%) | 7 (3%) | |
| 1,0,0 | 0 (0%) | 2 (1%) | |
| 1,0,1 | 2 (1%) | 2 (1%) | |
| 1,1,0 | 12 (5%) | 8 (4%) | |
| 1,1,1 | 230 (89%) | 190 (89%) | |
| Stent technique | 0.02 | ||
| Crush | 4 (2%) | 4 (2%) | |
| Stent main branch/angioplasty side branch | 174 (67%) | 118 (55%) | |
| T stent | 22 (9%) | 17 (8%) | |
| V stent | 38 (15%) | 57 (27%) | |
| Y stent | 20 (8%) | 17 (8%) | |
| Stent characteristics | |||
| Mean number of stents | 1.6 ± 0.7 | 1.8 ± 1.0 | 0.007 |
| Bare-metal stent(s) only | 35 (14%) | 33 (15%) | 0.55 |
| Drug-eluting stent(s) only | 219 (85%) | 158 (74%) | 0.004 |
| Combination bare-metal/drug-eluting stent(s) | 3 (1%) | 19 (9%) | <0.0001 |
| Mean stent length (mm) | 17.1 ± 5.2 | 15.0 ± 4.6 | <0.0001 |
IVUS findings observed during the index PCI are presented in Table 3 . IVUS was used in 34% of lesions (n = 87) before PCI and in all lesions after PCI. In almost 1/2 of bifurcation lesions, IVUS findings prompted further intervention. IVUS prompted a change in PCI strategy for 35 lesions (14%). Stent underexpansion was observed in 71 vessels (27%), prompting additional balloon dilation in all cases. Lesions with edge dissection (n = 7) or plaque shift (n = 14) also received further therapy. The remaining vessels (n = 131) were considered to have optimal stent dimensions and well-apposed stent struts. No significant differences were observed in clinical outcomes in patients who had IVUS performed before and after PCI compared to after PCI only.
| Variable | Total | Lesions Developing TLR |
|---|---|---|
| (n = 258 lesions) | (n = 15 lesions) | |
| Intravascular ultrasound used before stenting | 87 (34%) | 3 (15%) |
| Intravascular ultrasound used after stenting | 258 (100%) | 15 (100%) |
| Well-apposed stent/no complications | 131 (51%) | 8 (53%) |
| Poststenting intravascular ultrasound findings requiring intervention | 127 (49%) | 7 (46%) |
| Edge dissection | 7 (3%) | 0 (0%) |
| Lesion recharacterization requiring stenting | 35 (14%) | 2 (13%) |
| Plaque shift | 14 (5%) | 0 (0%) |
| Stent underexpansion | 71 (27%) | 5 (33%) |
Clinical follow-up was available for all patients with a mean follow-up period of 2.5 ± 2.0 years. Clinical outcomes for the study are presented in Table 4 . The composite rate of all-cause death or MI was significantly lower in the IVUS group compared to the no-IVUS group (8% vs 21%, p = 0.0002). Compared to no IVUS, IVUS use was associated with a significantly lower rate of death (6% vs 12%, p = 0.02), MI (3% vs 12%, p = 0.0002), stent thrombosis (0% vs 2%, p = 0.02), TVR (7% vs 24%, p <0.0001), and a trend toward a lower rate of periprocedural MI (6% vs 12%, p = 0.06). Occurrence of TLR was significantly less frequent in the IVUS group compared to the no-IVUS group, with an absolute difference of 15% (6% vs 21%, p <0.0001). In the IVUS group, TLR occurred evenly between the main branch and side branch, whereas main branch TLR was more frequent then side branch TLR in the no-IVUS group. Because of the larger number of patients with left main coronary artery disease in the IVUS (n = 37) versus no-IVUS (n = 4) group, we analyzed each clinical outcome excluding those patients and did not observe any significant change in overall results of the study.
