Intraoperative ticagrelor removal to reduce bleeding after urgent CABG: matched comparison of patient-level observational data

Background Patients on ticagrelor undergoing urgent CABG before completing the recommended 3-day washout are at high risk for severe perioperative bleeding. The Safe and Timely Antithrombotic Removal – Ticagrelor (STAR-T) randomized trial showed that the intraoperative use of a novel device capable of removing antithrombotic drugs is safe and reduces bleeding in patients undergoing CABG within 2 days of ticagrelor discontinuation. Purpose To evaluate whether real world use of the device reduces bleeding in patients on ticagrelor undergoing urgent CABG. Methods Bleeding after CABG <3 days from last ticagrelor dose was evaluated in 2 groups according to device use. Group 1 (CABG + Device, n=198) included patients from the international STAR registry, and Group 2 (CABG No Device, n=389) from a meta-analysis of 4 observational studies. Propensity scores were estimated using ten clinical and operative variables and inverse probability of treatment weighting (IPTW) was used to create balanced groups with target for absolute standardized mean difference (SMD) of <0.1. The primary outcome was Bleeding Academic Research Consortium (BARC)-4 bleeding. Results Results are reported for Group 1 followed by Group 2. After IPTW there were no significant differences in demographics or baseline characteristics between the two groups. Mean time between last ticagrelor dose and CABG was comparable (0.78 vs. 0.83 days; SMD: -0.06, p=0.45) as was median EuroSCORE-II (2.8% vs. 2.7%, SMD: -0.05, p=0.56) and median cardiopulmonary bypass duration (102 vs. 102 min, SMD: 0.01, p=0.81). The incidence of BARC-4 was significantly reduced by the device (21.8% vs. 34.5%; OR: 0.53; 95% CI: 0.35, 0.78; p=0.002). Sensitivity analysis using doubly robust estimation was consistent (OR: 0.53; 95% CI: 0.35, 0.79; p=0.002). Secondary outcomes analyses showed that the device significantly reduced 24-hour chest tube drainage (-125.4 mL; 95% CI: -244.2, -6.6, p=0.039) and red blood cell transfusions within 48 hours of CABG (-1.19 units; 95% CI: -1.97, -0.38, p=0.003). Re-operations for bleeding control (5.7% vs. 7.2%; OR: 0.77; p=0.47) and 30-day mortality (3.8% vs. 7.4%; p=0.12) were not significantly reduced by the device. Conclusion The current analysis showing significant reductions in BARC-4 bleeding in patients on ticagrelor undergoing urgent CABG validates the observations of the STAR-T randomized trial in the real-world setting. Therefore the device should be considered in patients undergoing urgent CABG under ticagrelor.