Platelet inhibitor withdrawal and outcomes after coronary artery surgery: an individual patient data meta-analysis

Abstract

OBJECTIVES To evaluate the association between guideline-conforming as compared to shorter than recommended withdrawal period of P2Y12 receptor inhibitors prior to isolated on-pump coronary artery bypass grafting (CABG) and the incidence of severe bleeding and ischaemic events. Randomized controlled trials are lacking in this field. METHODS We searched PUBMED, Embase and other suitable databases for studies including patients on P2Y12 receptor inhibitors undergoing isolated CABG and reporting bleeding and postoperative ischaemic events from 2013 to March 2024. The primary outcome was incidence of Bleeding Academic Research Consortium type 4 (BARC-4) bleeding defined as any of the following: perioperative intracranial bleeding, reoperation for bleeding, transfusion of ≥5 units of red blood cells, chest tube output of ≥2 l. The secondary outcome was postoperative ischaemic events according to the Academic Research Consortium 2 Consensus Document. Patient-level data provided by each observational trial were synthesized into a single dataset and analysed using a 2-stage IPD-MA. RESULTS Individual data of 4837 patients from 7 observational studies were synthesized. BARC-4 bleeding, 30-day mortality and postoperative ischaemic events occurred in 20%, 2.6% and 5.2% of patients. After adjusting for EuroSCORE II and cardiopulmonary bypass time, guideline-conforming withdrawal was associated with decreased BARC-4 bleeding risk in patients on clopidogrel [adjusted odds ratio (OR) 0.48; 95% confidence intervals (CI) 0.28–0.81; P = 0.006] and a trend towards decreased risk in patients on ticagrelor (adjusted OR 0.48; 95% CI 0.22–1.05; P = 0.067). Guideline-conforming withdrawal was not significantly associated with 30-day mortality risk (clopidogrel: adjusted OR 0.70; 95% CI 0.30–1.61; ticagrelor: adjusted OR 0.89; 95% CI 0.37–2.18) but with decreased risk of postoperative ischaemic events in patients on clopidogrel (clopidogrel: adjusted OR 0.50; 95% CI 0.30–0.82; ticagrelor: adjusted OR 0.78; 95% CI 0.45–1.37). BARC-4 bleeding was associated with 30-day mortality risk (adjusted OR 4.76; 95% CI 2.67–8.47; P < 0.001). CONCLUSIONS Guideline-conforming preoperative withdrawal of ticagrelor and clopidogrel was associated with a 50% reduced BARC-4 bleeding risk when corrected for EuroSCORE II and cardiopulmonary bypass time but was not associated with increased risk of 30-day mortality or postoperative ischaemic events.