SYNTAX
PubMed • Full text • PDF
Clinical Question
Among patients with 3-vessel and/or left main disease, how does PCI compare with CABG in terms of rates of major CV events?
Bottom Line
CABG resulted in fewer major CV events at 1 year compared with PCI among patients with 3-vessel and/or left main disease.
Major Points
The Synergy between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery (SYNTAX) Trial randomized 1,800 patients with 3-vessel and/or left main CAD to either PCI with paclitaxel-eluting stents or CABG. SYNTAX demonstrated that major CV events occurred more frequently in the PCI arm (17.8% vs. 12.4%), which was driven primarily by the increased frequency of repeat revascularization procedures in patients treated with PCI (13.5% vs. 5.9%). However, there were more strokes in the CABG arm (0.6% vs. 2.2%).
SYNTAX developed a score to classify patients according to severity of their CAD by incorporating lesion complexity, location, and number. In the SYNTAX trial, patients with lower SYNTAX scores had lower event rates than those with higher scores. Secondary analyses suggest that among patients with low SYNTAX scores, outcomes were similar between the two arms, whereas among patients with high SYNTAX scores, outcomes were better with CABG.
Guidelines
Adapted from ACCF/AHA/ACP/AATS/PCNA/SCAI/STS Stable Ischemic Heart Disease (2012)[1]
- Left main disease with ≥50% diameter stenosis, for survival
- CABG recommended (class I level B)
- PCI is reasonable if both of the following (class IIa level B):
- Anatomically low risk of PCI complications and high likelihood of good long-term outcome exemplified by SYNTAX score ≤22, ostial, or trunk left main disease
- Increased risk of adverse surgical outcomes exemplified by STS risk score ≥5%
- PCI may be reasonable if both of the following (class IIb level B):
- Anatomically low to intermediate risk of PCI complications and intermediate to high likelihood of good long-term outcome exemplified by SYNTAX score ≤33 and bifurcation left main disease
- Increased risk of adverse surgical outcomes exemplified by STS risk score ≥2%, moderate or severe COPD, prior stroke with disability, or prior cardiac surgery
- PCI shouldn't be performed in patients with ≥50% diameter stenosis of left main if unfavorable PCI anatomy or if they are good candidates for CABG (class III level B)
- Non-left main CAD, for survival
- CABG recommended if ≥70% stenosis in ≥3 major coronary arteries, with or without proximal LAD involvement (class I level B)
- CABG recommended if ≥70% stenosis in proximal LAD plus 1 other major coronary artery (class I level B)
- CABG reasonable if >70% stenosis in 2 major coronary arteries with target vessels supplying a large area of viable tissue or severe/extensive ischemia exemplified by high risk stress testing, abnormal intracoronary hemodynamics on evaluation, or >20% perfusion defect on stress imaging (class IIa level B)
- CABG reasonable if EF 35-50% and ≥70% multivessel or proximal LAD stenosis if viable tissue is present in area downstream from intended lesions (class IIa level B)
- CABG via LIMA if ≥70% left main disease and extensive ischemia (class IIa level B)
- CABG probably better than PCI if DM and multivessel CAD -- especially if LIMA to LAD anastomosis can be performed (class IIa level B)
- No CABG or PCI stable for stable ischemic heart disease if non-significant stenoses (<70% diameter, FFR>0.80, mild ischemia on noninvasive testing), only left circumflex or right coronary, or perfuse only small area of tissue (class III level B)
Design
- Multicenter, parallel-group, randomized, controlled trial
- N=1,800 patients with 3-vessel and/or left main CAD
- PCI with paclitaxel-eluting stents (n=903)
- CABG (n=897)
- Setting: 85 centers across 17 countries in Europe and US
- Enrollment: 2005-2007
- Median follow-up: 1 year
- Analysis: Intention-to-treat, noninferiority
- Primary outcome: major CV event by 12 mos
Population
Inclusion Criteria
- ≥50% stenosis in ≥3 coronary arteries and/or left main CAD
- Both treatment options (CABG or PCI) could achieve equivalent revascularization as judged by an interventional cardiologist and cardiac surgeon
Exclusion Criteria
- Prior PCI or CABG
- Acute MI
- Only one treatment option was suitable for revascularization
- 1,077 enrolled in CABG registry; 198 enrolled in PCI registry
Baseline Characteristics
- Demographics: Age 65 yrs, male 78%
- Baseline health data: BMI 28kg/m2
- PMH: Medically treated DM: 25.1% (of which 40.2% required insulin), metabolic syndrome 45.8%, current smoker 20.2%, CAD 8.2%, MI 32.8%, previous stroke/TIA 9.0%, stable angina 57.1%, unstable angina 28.5%
- Vital signs: BP ≥130/85mmHg: 68.9% vs. 64.0% (P=0.03)
- CHF: 4.6%; EF <30%: 1.3% vs. 2.5% (P=0.08)
- Labs: HL 78%, TG ≥150mg/dl 32.3% vs. 38.7% (P=0.007), HDL <40mg/dl for men or <50mg/dl for women 46.2% vs. 52.5% (P=0.01)
- euroSCORE: 3.8
- Parsonnet score: 8.5
- SYNTAX score: 28.7
- Number of lesions: 4.3
- Total occlusion: 23.2%
- Bifurcation: 72.8%
- Time to procedure: 6.9 vs. 17.4 days (P<0.001)
- Procedure duration: 1.7 vs. 3.4 hrs (P<0.001)
- Postprocedural hospital stay: 3.4 vs. 9.5 days (P<0.001)
- Complete revascularization: 56.7% vs. 63.2% (P=0.005)
Interventions
- Patients with 3-vessel and/or left main CAD determined by a local interventional cardiologist and cardiac surgeon at each site to be eligible for both PCI or CABG were randomized to PCI with paclitaxel-eluting stents or CABG
- Goal of therapy was to achieve complete revascularization of all target vessels
- Patients for whom only one treatment option was suitable were entered into parallel, nested registry: PCI registry for CABG-ineligible and CABG registry for PCI-ineligible
Post-Procedure Medications
Comparisons are PCI vs. CABG. All P-values are <0.001 unless otherwise stated.
- Aspirin: 96.3% vs. 88.5% at discharge; 91.2% vs. 84.3% at 12 mos
- Thienopyridine: 96.8% vs. 19.5% at discharge; 71.1% vs. 15.0% at 12 mos
- Warfarin derivative: 2.6% vs. 7.1%
- Statin: 86.7% vs. 74.5%
- Beta-blocker: 81.3% vs. 78.6% (P=0.17)
- ACE inhibitor: 55.1% vs. 44.6%
- Calcium-channel blocker: 25.8% vs. 18.4%
- ARB: 13.3% vs. 7.0%
- Amiodarone: 1.5% vs. 12.8%
Outcomes
Comparisons are PCI vs. CABG.
Primary Outcomes
- Major CV event (including death, stroke, MI, repeat revascularization) at one year
- 17.8% vs. 12.4% (RR 1.44; 95% CI 1.15-1.81; P=0.002; NNT=19)
Secondary Outcomes
- Death, stroke or MI
- 7.6% vs. 7.7% (RR 1.00; 95% CI 0.72-1.38; P=0.98)
- Death: 4.4% vs. 3.5% (P=0.37)
- Stroke: 0.6% vs. 2.2% (P=0.003)
- MI: 4.8% vs. 3.3% (P=0.11)
- Repeat revascularization
- 13.5% vs. 5.9% (RR 2.29; 95% CI 1.67-3.14; P<0.001)
- Graft occlusion or stent thrombosis
- 3.3% vs. 3.4% (P=0.89)
- Early graft occlusion or stent thrombosis: 2.0% vs. 0.3% (P=0.001)
- Late graft occlusion or stent thrombosis: 1.0% vs. 2.5% (P=0.02)
Subgroup Analysis
- Major CV event by 12 mos
- Among low SYNTAX score: 13.6 vs. 14.7 (P=0.71)
- Among intermediate SYNTAX score: 16.7 vs. 12.0 (P=0.10)
- Among high SYNTAX score: 23.4 vs. 10.9 (P<0.001)
Criticisms
- Short duration of follow-up. Follow-up at 4 years suggested diverging death rates in favor of CABG in patients with 3VD and LM CAD (Paul Taggart, CABG in 2012. Presented at the Southern Thoracic Surgical Association, 2011.)
- Male predominance (78%)
- Suboptimal use of medical therapy in CABG cohort, which may have led to more strokes
- Use of only paclitaxel stents, which have higher rate of angiographic restenosis than either the sirolimus-or everolimus-eluting stents
- Need for more pharmacologic support after PCI. Cessation of dual anti-platelet therapy for ASA mono-therapy after one year yielded increased risk of MACE in patients with DES (REAL LATE study).
Funding
Supported by Boston Scientific. Authors with multiple disclosures.