摘要：Summary Background On-pump cardiac surgery provokes a predictable perioperative myocardial ischaemia–reperfusion injury which is associated with poor clinical outcomes. We determined the occurrence of time-of-the-day variation in perioperative myocardial injury in patients undergoing aortic valve replacement and its molecular mechanisms. Methods We studied the incidence of major adverse cardiac events in a prospective observational single-centre cohort study of patients with severe aortic stenosis and preserved left ventricular ejection fraction (>50%) who were referred to our cardiovascular surgery department at Lille University Hospital (Lille, France) for aortic valve replacement and underwent surgery in the morning or afternoon. Patients were matched into pairs by propensity score. We also did a randomised study, in which we evaluated perioperative myocardial injury and myocardial samples of patients randomly assigned (1:1) via permuted block randomisation (block size of eight) to undergo isolated aortic valve replacement surgery either in the morning or afternoon. We also evaluated human and rodent myocardium in ex-vivo hypoxia–reoxygenation models and did a transcriptomic analysis in myocardial samples from the randomised patients to identify the signalling pathway(s) involved. The primary objective of the study was to assess whether myocardial tolerance of ischaemia–reperfusion differed depending on the timing of aortic valve replacement surgery (morning vs afternoon), as measured by the occurrence of major adverse cardiovascular events (cardiovascular death, myocardial infarction, and admission to hospital for acute heart failure). The randomised study is registered with ClinicalTrials.gov, number NCT02812901. Findings In the cohort study (n=596 patients in matched pairs who underwent either morning surgery [n=298] or afternoon surgery [n=298]), during the 500 days following aortic valve replacement, the incidence of major adverse cardiac events was lower in the afternoon surgery group than in the morning group: hazard ratio 0·50 (95% CI 0·32–0·77; p=0·0021). In the randomised study, 88 patients were randomly assigned to undergo surgery in the morning (n=44) or afternoon (n=44); perioperative myocardial injury assessed with the geometric mean of perioperative cardiac troponin T release was significantly lower in the afternoon group than in the morning group (estimated ratio of geometric means for afternoon to morning of 0·79 [95% CI 0·68–0·93; p=0·0045]). Ex-vivo analysis of human myocardium revealed an intrinsic morning–afternoon variation in hypoxia–reoxygenation tolerance, concomitant with transcriptional alterations in circadian gene expression with the nuclear receptor Rev-Erbα being highest in the morning. In a mouse Langendorff model of hypoxia–reoxygenation myocardial injury, Rev-Erbα gene deletion or antagonist treatment reduced injury at the time of sleep-to-wake transition, through an increase in the expression of the ischaemia–reperfusion injury modulator CDKN1a/p21. Interpretation Perioperative myocardial injury is transcriptionally orchestrated by the circadian clock in patients undergoing aortic valve replacement, and Rev-Erbα antagonism seems to be a pharmacological strategy for cardioprotection. Afternoon surgery might provide perioperative myocardial protection and lead to improved patient outcomes compared with morning surgery. Funding Fondation de France, Fédération Française de Cardiologie, EU-FP7-Eurhythdia, Agence Nationale pour la Recherche ANR-10-LABX-46, and CPER-Centre Transdisciplinaire de Recherche sur la Longévité.