Randomized Clinical Trial of Surgical vs. Percutaneous vs. Hybrid Revascularization in Multivessel Coronary Artery Disease: Residual Myocardial Ischemia and Clinical Outcomes at One Year-Hybrid coronary REvascularization Versus Stenting or Surgery (HREVS)

Abstract

Aim: Optimal revascularization strategy in multivessel (MV) coronary artery disease (CAD) eligible for percutaneous management (PCI) and surgery remains unresolved. We evaluated, in a randomized clinical trial, residual myocardial ischemia (RI) and clinical outcomes of MV-CAD revascularization using coronary artery bypass grafting (CABG), hybrid coronary revascularization (HCR), or MV-PCI.

Methods: Consecutive MV-CAD patients (n = 155) were randomized (1 : 1 : 1) to conventional CABG (LIMA-LAD plus venous grafts) or HCR (MIDCAB LIMA-LAD followed by PCI for remaining vessels) or MV-PCI (everolimus-eluting CoCr stents) under Heart Team agreement on equal technical and clinical feasibility of each strategy. SPECT at 12 months (primary endpoint of RI that the trial was powered for; a measure of revascularization midterm efficacy and an independent predictor of long-term prognosis) preceded routine angiographic control.

Results: Data are given, respectively, for the CABG, HCR, and MV-PCI arms. Incomplete revascularization rate was 8.0% vs. 7.7% vs. 5.7% (p=0.71). Hospital stay was 13.8 vs. 13.5 vs. 4.5 days (p < 0.001), and sick-leave duration was 23 vs. 16 vs. 8 weeks (p < 0.001). At 12 months, RI was 5 (2, 9)% vs. 5 (3, 7)% vs. 6 (3, 10)% (median; Q1, Q3) with noninferiority p values of 0.0006 (HCR vs. CABG) and 0.016 (MV-PCI vs. CABG). Rates of angiographic graft stenosis/occlusion or in-segment restenosis were 20.4% vs. 8.2% vs. 5.9% (p=0.05). Clinical target vessel/graft failure occurred in 12.0% vs. 11.5% vs. 11.3% (p=0.62). Major adverse cardiac and cerebral event (MACCE) rate was similar (12% vs. 13.4% vs. 13.2%; p=0.83).

Conclusion: In this first randomized controlled study comparing CABG, HCR, and MV-PCI, residual myocardial ischemia and MACCE were similar at 12 months. There was no midterm indication of any added value of HCR. Hospital stay and sick-leave duration were shortest with MV-PCI. While longer-term follow-up is warranted, these findings may impact patient and physician choices and healthcare resources utilization. This trial is registered with NCT01699048.

Figure 1

Noninferiority analysis for the SPECT-based residual ischemia at 12 months in the three treatment arms with CABG as a reference method (a) and assuming no single reference method (b). Point estimates and 90% confidence intervals for the differences in RI between treatment modalities are shown with solid vertical gridline indicating the null difference and interrupted vertical gridline indicating the noninferiority margin of 4.2 percentage points. (a) Respective p values are for noninferiority of MV-PCI vs CABG and HCR vs. CABG. To adjust for two comparisons with CABG as the reference p values were considered statistically significant when <0.025. (b) p values are for pairwise noninferiority tests with 95% one-sided confidence intervals. Overall p for noninferiority is 0.046.

Figure 2

MACCE-free survival (a) and cumulative risk of MACCE (b) during 12-month follow-up according to the treatment arm. Panel A shows MACCE-free survival, whereas the cumulative risk of MACCE is depicted in Panel B. Numbers of patients at risk are shown above the horizontal axis in panel A. Pairwise comparisons of treatment arms with Cox proportional hazards model are shown at the bottom of panel B. MACCE—Major Adverse Cardiac or Cerebral Event.