非急性心肌缺血综合征的血运重建策略：贝叶斯网络荟萃分析

原文

BACKGROUND: Contemporary guidelines by the European Society for Cardiology and American College of Cardiology/American Heart Association for the treatment of non-acute myocardial ischaemic syndromes dispute the value of revascularisation and differ in their recommendation to perform revascularisation. A Bayesian network meta-analysis was performed, evaluating the strength of evidence for the comparative incremental effectiveness of coronary artery bypass grafting (CABG) versus percutaneous coronary intervention (PCI) over medical therapy on long-term outcomes. METHODS: A hierarchical Bayesian network meta-analysis was designed (PROSPERO CRD42024541215, date 20 May 2024), including randomised controlled trials (RCTs) published between 2005 and 10 June 2025, which consisted of three initial treatment modalities: optimal medical therapy (OMT), PCI+OMT and CABG+OMT. The primary outcome was all-cause mortality at maximum follow-up; secondary outcomes were trates of the rates of myocardial infarction, stroke and re-revascularisation at maximum follow-up, expressed in HRs and 95% credible intervals (CrIs), accompanied by surface under the cumulative ranking curve (SUCRA) scores. RESULTS: 10 RCTs, comprising 10 742 patients, were included. For all-cause mortality, the estimated median HR of CABG+OMT versus OMT was 0.84 (95% CrI 0.68-1.07); the HR of PCI+OMT versus OMT was 0.93 (0.79-1.16); and the HR of CABG+OMT versus PCI+OMT was 0.91 (0.71-1.13). The SUCRAs of a CABG+OMT strategy ranking as the optimal revascularisation treatment regarding mortality, myocardial infarction, stroke and re-revascularisation were 88.1%, 99.7%, 17.5% and 99.5%, respectively. Results were consistent across sensitivity analyses, including in the node-splitting models. CONCLUSIONS: This Bayesian network meta-analysis found that an initial CABG (+OMT) revascularisation strategy was associated with higher probabilities of optimal outcomes, with the exception of stroke, compared with an initial PCI (+OMT) revascularisation strategy, although CrIs overlapped, suggesting that some uncertainty remains. PROSPERO REGISTRATION NUMBER: CRD42024541215.