Use of coronary artery bypass graft surgery and percutaneous coronary intervention and associated outcomes in the ISCHEMIA trial.

BACKGROUND: In the ISCHEMIA Trial, 5,179 patients with stable coronary disease were randomized to initial invasive or conservative management. METHODS: PCI was recommended with a SYNTAX score 0 to 22 (low) and CABG with a SYNTAX score ≥33 (high). Either could be recommended for intermediate scores. The composite primary outcome was cardiovascular death, MI, or hospitalization for unstable angina, heart failure, or resuscitated cardiac arrest. There were 2 cohorts in this analysis. The descriptive cohort included patients who underwent CABG or PCI within 180 days of randomization and had no primary outcome before revascularization. The comparative cohort excluded patients with prior CABG, single vessel disease, SYNTAX score ≥ 45, and without core laboratory assessment. We focused on the intermediate (23-32) SYNTAX comparative group for which either CABG or PCI could be recommended. RESULTS: For 1,935 patients in the descriptive cohort (485 CABG, 1,450 PCI), the SYNTAX score was 27.3 ± 11.0 in the CABG group and 15.3 ± 8.6 in the PCI group, P < .0001. Most patients with low SYNTAX scores underwent PCI (87.1%), while most with high SYNTAX scores underwent CABG (72.6%). For the 1,203 patients (385 CABG, 818 PCI) in the comparative cohort, the adjusted 4-year primary event rate was 14.5% for CABG and 13.2% for PCI (difference 1.3%, 95% CI, -4.9% to 7.7%). For the 346 patients (163 CABG, 183 PCI) in the intermediate SYNTAX group, the adjusted 4-year primary event rate was 10.6% for CABG and 18.3% for PCI (difference -7.6%, 95% CI, -16.1% to 0.9%). CONCLUSIONS: Selection of revascularization method resulted in more PCI in the low SYNTAX group and more CABG in the high SYNTAX group. There was no statistical evidence of a difference between PCI and CABG in the intermediate SYNTAX group but the CIs are broad, reflecting uncertainty. GOV IDENTIFIER: NCT01471522; https://clinicaltrials.gov/ct2/show/NCT01471522.

Duke Scholars

Author Daniel Benjamin Mark Medicine, Cardiology

Author Sean Michael O'Brien Biostatistics & Bioinformatics, Division of Biostatistics