A Comparison of the Rates of Gastrointestinal Bleeding and Mortality in Cirrhotic Patients with Coronary Artery Disease Treated with Cardiac Stents vs Medical Management

Abstract

Introduction: The reported prevalence of coronary artery disease (CAD) in patients with cirrhosis is as high as 28%. Since cirrhotic patients are at risk for GI bleeding (GIB) due to varices and coagulopathy, it is not clear if the benefits of treating their CAD with coronary stents outweigh the risks of GIB from the dual antiplatelet therapy (DAPT) required to prevent stent thrombosis. To address this issue, we assessed the rates of mortality and GIB in cirrhotic patients whose CAD was treated with cardiac stenting or with medical management.Table 1Table 2Methods: We conducted a retrospective review of medical records to identify all patients with: 1) CAD diagnosed by coronary angiography or stress test, and 2) cirrhosis diagnosed clinically or radiographically between January 2000 and September 2015. Patients were divided into two groups based on CAD treatment (stent or medical), and outcomes were compared between groups. Primary outcomes studied were rates of GIB (evaluated at 30 days, 90 days, 1 year, and 2 years) and mortality (evaluated at 30 days, 90 days, and 1 year). Results: We identified 148 patients with both CAD and cirrhosis (mean age 60.4 ± 0.6 years, 98% men); 68 had coronary stent implantation (cases) and 80 were managed medically (controls). Cases and controls did not differ significantly in basic demographics, Charlson Comorbidity Index, MELD scores, or indication for cardiac work-up (see table); the frequency of hyperlipidemia and prior myocardial infarction were significantly higher in cases, and the frequency of combined hepatitis C and alcohol was higher in controls. DAPT was used for ≥ 30 days in 99% of cases, and in only 5% of controls. The incidence of GIB was 7.4% vs. 1.3% at 30days (p=0.09), 10.3% vs. 2.5% at 90days (p=0.08), 22.1% vs. 5% at 1year (p=0.003) and 27.9% vs. 5% at 2 year (p=0.0002). Interestingly, there was no significant difference between the groups in mortality at any time point (cases vs. controls: 2.9% vs. 1.3% at 30 days, 2.9% vs. 3.8% at 90 days, and 11.8% vs. 13.8% at 1 year). Moreover, rates of GIB were similar among patients who received DES and BMS (26.9% vs. 31.3%, p=0.76). Conclusion: For our patients who had both cirrhosis and CAD, those treated with coronary stents had a significantly higher rate of GIB, but no significantly different rate of mortality, than those whose CAD was treated medically. This information is useful when determining the risk/benefit ratio of PCI among patients with cirrhosis.