Abstract 11232: Coronary Artery Bypass Graft Surgery Shows Improved Restricted Mean Survival Time and Time Lost Compared to Medical Therapy in Patients With Left Ventricular Dysfunction: From the STICH Trial

Background: Restricted mean survival time (RMST) and restricted mean time lost (RMTL) are alternative clinical trial outcome measures that can be used to quantify treatment effects. In this study we evaluated RMST and RMTL in the STICH trial to quantify long-term treatment effects of coronary artery bypass graft surgery (CABG) vs. medical therapy in patients with coronary artery disease (CAD) and depressed left ventricular ejection fraction (LVEF). Methods: RMST was calculated as the average event-free survival up to follow-up at 10 years. RMST was calculated separately using i) the trial follow-up time and ii) age instead of time with estimation of age-specific events rates. Delta RMST was determined for CABG vs. medical therapy, with delta RMST >0 favoring CABG. The RMTL ratio was calculated by dividing the RMTL of the CABG group by the RMTL of the medical therapy group, with a ratio <1 favoring CABG. Analysis was performed for the primary and secondary outcomes for the entire cohort and prespecified subgroups. Results: For the primary outcome of all-cause mortality, RMST for the CABG group was 2714 days vs. 2501 days for the medical therapy group (delta RMST = 213 days [95% CI 21.63, 404.63]; p=0.029). The RMTL ratio for all-cause mortality was 0.90 (95% CI 0.82, 0.99; p = 0.030). For the secondary outcomes of death from cardiovascular (CV) causes, all-cause mortality or CV hospitalization, and all-cause mortality or revascularization, delta RMST and the RMTL ratio favored CABG over medical therapy (Figure). In subgroup analyses of all-cause death, delta RMST and RMTL ratio favored CABG in patients under 60 years old, racial/ethnic minorities, and those with 3-vessel disease. Conclusion: In 10-year follow up in the STICH trial, RMST and RMTL both favored CABG compared to medical therapy for all-cause mortality and other cardiovascular endpoints in patients with CAD and low LVEF. These findings can be used to help quantify the potential benefits of CABG in this population.