Upper gastrointestinal bleeding in patients with stable coronary artery disease (registry of antithrombotic therapy "REGATТA" results)

Abstract

Upper gastrointestinal (UGI) bleeding is a common complication of antiplatelet therapy. Data from real clinical practice that characterize the range of risk factors for UGI bleeding, prophylactic proton pump inhibitors (PPIs) therapy, bleeding frequency and their long-term effects in patients with stable coronary artery disease (CAD) are limited. To identify predictors of UGI bleeding in patients with stable CAD, to assess the role of PPI in the prevention of bleeding and the long-term prognosis of patients after bleeding. 934 patients with stable CAD (median age 61 [5368] years, 78.6% men) were included in the single institution prospective REGistry of Long-term AnTithrombotic TherApy (REGATTA). Atherosclerosis of peripheral arteries (APA) and abdominal aortic aneurysm (AAA) screening was performed by doctor decision, as well as esophagogastroduodenoscopy. 76% of patients received dual antiplatelet therapy for 612 months after elective PCI. PPIs were prescribed in 28.3% of cases. The median follow-up was 2.5 [1.15.1] years. The frequency of overt UGI bleeding was 1.9 per 100 patients per year. Anamnesis of peptic ulcer disease (OR 4.7; 95% CI 1.911.8;p=0.001), erosion of the upper gastrointestinal tract (OR 6.7; 2.716.6;p=0.00004 ), as well as concomitant diseases associated with a decrease in blood supply to the mucosa, such as heart failure HF (OR 6.1; 2.316.0;p=0.0002), AAA (OR 9.3; 2.534.2;p=0.0008) and APA (OR 2.3; 0.985.5;p=0.05) turned out to be independent predictors of UGI bleeding. The frequency of AAA among those who underwent UGI bleeding was 19.6% (in patients without bleeding 1.4%;p0.001). 90.2% of patients with UGI bleeding received PPI; the frequency of UGI bleeding in patients receiving pantoprazole and omeprazole did not differ significantly. After UGI bleeding, rebleeding rate was 7.8%, thrombotic events (TE) rate 31.4%, mortality rate 17.7% for 30 days, 19.4% for 1 year and 35.3% for the entire observation period. The predictors of deaths were AAA (OR 92.5; 7.7107.9;p0.0001), APA (OR 4.2; 1.0317.2;p=0.045) and HF (OR 34.5; 8.5140.6;p0.0001). The worst prognosis was expected for patients who underwent UGI bleeding and thrombotic events: 2/3 of these patients died. In a prospective analysis of patients with stable CAD, we identified UGI bleeding was a significant risk factor for late thromboembolism and death, compared with patients without bleeding. Predictors of UGI bleeding and poor prognosis are factors that indicate atherothrombotic burden abdominal aortic aneurysm, peripheral atherosclerosis and HF. Registration: URL: https://www.clinicaltrials.gov; Unique identifier: NCT04347200.