PERIPROCEDURAL BLEEDING IS A STRONG INDEPENDENT PREDICTOR OF LONG-TERM PROGNOSIS AFTER ELECTIVE PERCUTANEOUS CORONARY INTERVENTION. LONG-TERM FOLLOW-UP ANALYSIS OF THE PROSPECTIVE, MULTICENTER, RANDOMIZED PRAGUE-8 TRIAL

Background: The aim of presented analysis was to investigate the impact of procedure-related bleeding on long-term outcomes after elective PCI and to assess the appropriateness, including the risk of periprocedural bleeding, in decision-making process about invasive/medical management for stable CAD.
Methods: In the PRAGUE-8 trial, 1028 patients were randomly assigned: to "nonselective" - clopidogrel 600 mg >6 hours before elective coronary angiography, or to "selective" - clopidogrel 600 mg in the cath-lab after coronary angiography, only in case of PCI. The follow-up period (2.7 ± 0.91 years) was completed in 99.3% of patients who underwent PCI in this study. Bleeding - defined according to the TIMI criteria - included all procedure-related bleeding events. The primary outcome was long-term mortality.
Results: 4% of patients experienced a PCI-related bleeding. During the follow-up investigation, the incidence of all-cause mortality associated with a bleeding event was 7.39% (in a group of 284 patients without bleeding) versus 25% (in a group of 12 patients with PCI-related bleeding complications) (OR 4.17, 95% C.I. 1.05 to 16.60, p=0.042). The Kaplan-Meier estimate of mortality at 3 years was 33.3% among patients with periprocedural bleeding versus 7.7% among patients without bleeding complication. A Cox proportional hazards model revealed that periprocedural bleeding (HR 5.62, 95% C.I. 1.40 to 22.58; p = 0.015) independently predicted long-term mortality after adjustment for age, gender, DM, renal failure, hypertension and NYHA.
Conclusion: Procedure-related bleeding is a strong independent predictor of long-term mortality after elective PCI. Data from the randomized PRAGUE-8 study underscores the importance of including risk of procedure-related bleeding in the decision-making process regarding management strategies for stable CAD.